article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the dirac phase factor @xmath2 with a phase containing spatial or temporal integrals of potentials ( of the general form @xmath3 ) @xmath2 is the standard and widely used solution of the gauge transformation equations of electrodynamics ( with @xmath4 and @xmath5 vector and scalar potentials respectively ) . in a quantum mechanical context
, it connects wavefunctions of two systems ( with different potentials ) that experience the same classical fields at the observation point @xmath6 , the two more frequently discussed cases being : either systems that are completely gauge - equivalent ( a trivial case with no physical consequences ) , or systems that exhibit phenomena of the aharonov - bohm type ( magnetic or electric)@xcite @xmath2 and then this dirac phase has nontrivial observable consequences ( mathematically , this being due to the fact that the corresponding gauge function is now multiple - valued ) . in the above two cases , the classical fields experienced by the two ( mapped ) systems are equal _ at every point _ of the accessible spacetime region .
however , it has not been widely realized that the gauge transformation equations , viewed in a more general context , can have _ more general solutions _ than simple dirac phases , and these lead to wavefunction-_phase - nonlocalities _ that have been widely overlooked and that seem to have important physical consequences .
these nonlocal solutions are applicable to cases where the two systems are allowed to experience _ different fields _ at spacetime points ( or regions ) that are _ remote _ to ( and do _ not _ contain ) the observation point @xmath6 ( these regions being _ physically accessible _ to the particle , unlike genuine aharonov - bohm cases ) . in this article
we rigorously show the existence of these generalized solutions , demonstrate them in simple physical examples , and fully explore their structure , presenting cases ( and closed analytical results for the wavefunction - phases ) that actually connect ( or map ) two quantal systems that are * neither physically equivalent nor of the usual aharonov - bohm type*. we also fully investigate the consequences of these generalized ( _ nonlocal _ ) influences ( on wavefunction - phases ) and find them to be numerous and important ; we actually find them to be of a different type in static and in time - dependent field - configurations ( and in the latter cases we show that they lead to relativistically _ causal _ behaviors , that apparently resolve earlier paradoxes arising in the literature from the use of standard dirac phase factors ) . the nonlocal phase behaviors discussed in the present work may be viewed as a justification for the ( recently emphasized@xcite ) terminology of dynamical nonlocalities associated with all aharonov - bohm effects ( even static ones ) , although in our approach these nonlocalities seem to also respect causality ( without the need to independently invoke the uncertainty principle ) and , to the best of our knowledge , this is the first theoretical picture with such characteristics . in order to introduce some background and further motivation for
this article let us first remind the reader of a very basic property that will be central to everything that follows , which however is usually taken to be valid only in a restricted context ( but is actually more general than often realized ) .
this property is a simple ( u(1 ) ) phase - mapping between quantum systems , and is usually taken in the context of gauge transformations , ordinary or singular@xmath7 here , however , it will appear in a more general framework , hence the importance of reminding of its independent , basic and more general origin .
we begin by recalling that , if @xmath8 and @xmath9 are solutions of the time - dependent schrdinger ( or dirac ) equation for a quantum particle of charge @xmath10 that moves ( as a test particle ) in two distinct sets of ( predetermined and classical ) vector and scalar potentials ( @xmath11 ) and ( @xmath12 ) , that are generally spatially- and temporally - dependent [ and such that , at the spacetime point of observation @xmath13 , the magnetic and electric fields are the same in the two systems ] , then we have the following formal connection between the solutions ( wavefunctions ) of the two systems @xmath14 with the function @xmath15 required to satisfy @xmath16 the above property can be immediately proven by substituting each @xmath17 into its corresponding ( @xmath18 ) time - dependent schrdinger equation ( namely with the set of potentials ( @xmath19))@xmath1 one can then easily see that ( [ basic1 ] ) and ( [ basic11 ] ) guarantee that both schrdinger equations are indeed satisfied together ( after cancellation of a few terms and then elimination of a global phase factor in system 2 ) . [ in addition , the equality of all classical fields at the observation point , namely @xmath20 @xmath21 for the magnetic fields and @xmath22 for the electric fields , is obviously consistent with all equations ( [ basic11 ] ) ( as is easy to see if we take the _ curl _ of the 1st and the _ grad _ of the 2nd ) @xmath2 provided , at least , that @xmath15 is such that interchanges of partial derivatives with respect to all spatial and temporal variables ( at the point @xmath23 ) are allowed ] .
as already mentioned , the above fact is of course well - known within the framework of the theory of quantum mechanical gauge transformations ( the usual case being for @xmath24 , hence for a mapping from a system with no potentials)@xmath7 but in that framework , these transformations are supposed to connect ( or map ) two _ physically equivalent systems _
( more rigorously , this being true for ordinary gauge transformations , in which case the function @xmath15 , the so - called gauge function , is unique ( single - valued ) in spacetime coordinates ) . in a formally similar manner , the above argument is also often used in the context of the so - called singular gauge transformations , where @xmath25 is multiple - valued , but the above equality of classical fields is still imposed ( at the observation point , which always lies in a physically accessible region)@xmath7 then the above simple phase mapping ( at all points of the physically accessible spacetime region , that always and everywhere experience equal fields ) leads to the standard phenomena of the aharonov - bohm type , where _ unequal fields in physically inaccessible regions _ have observable consequences .
however , we should keep in mind that that above property ( ( [ basic1 ] ) and ( [ basic11 ] ) taken together ) can be _ more generally valid _ @xmath2 and in this article we will present cases ( and closed analytical results for the appropriate phase connection @xmath15 ) that actually connect ( or map ) two systems ( in the sense of ( [ basic1 ] ) ) that are _ neither physically equivalent nor of the usual aharonov - bohm type_. and naturally , because of the above provision of field equalities at the observation point , it will turn out that any nonequivalence of the two systems will involve _ remote _ ( although _ physically accessible _ ) regions of spacetime , namely regions that do _ not _ contain the observation point @xmath13 ( and in which regions , as we shall see , the classical fields experienced by the particle may be _ different _ in the two systems ) .
one may wonder on the actual reasons why one should be looking for more general cases of a simple phase mapping of the type ( [ basic1 ] ) between _ nonequivalent _ systems . to answer this ,
let us take a step back and first recall some simple and well - known results that originate from the above phase mapping .
it is standard knowledge , for example , that , if we want to find solutions @xmath26 of the @xmath27-dependent schrdinger ( or dirac ) equation for a quantum particle ( of charge @xmath28 that moves along a ( generally curved ) one - dimensional ( 1-d ) path , and in the presence ( somewhere in the embedding 3-dimensional ( 3-d ) space ) of a fairly localized ( and time - independent ) classical magnetic flux @xmath29 that _ does * not * pass through any point of the path , _ then we formally have @xmath30 ( the dummy variable @xmath31 describing points along the 1-d path , and the term formally signifying that the above is valid _ before imposition of any boundary conditions _ ( meaning that these are to be imposed only on @xmath32 and not necessarily on @xmath33 ) . in ( [ standard ] ) , @xmath34 is a formal solution of the same system in the case of absence of any potentials ( hence with magnetic flux @xmath35 everywhere in the 3-d space ) .
the above holds because , for _ all points _ @xmath31 of the 1-d path , the particle experiences a vector potential @xmath36 of the form @xmath37 ( since the magnetic field is @xmath38 * * for _ all _ @xmath31 , by assumption ) , in combination with the above phase - mapping ( with a phase @xmath39 ) between two quantum systems , one in the presence and one in the absence of a vector potential ( i.e. the potentials in ( [ basic11 ] ) being @xmath40 and @xmath41 @xmath42 , together with @xmath43 if we decide to attribute everything to vector potentials only ) . in this particular system , the obvious @xmath44 that solves the above @xmath45 ( for * all * points of the 1-d space available to the particle ) is indeed @xmath46 ( since @xmath47 ) , and this gives ( [ standard ] ) ( if @xmath48 denotes the above point @xmath49 of observation and @xmath50 the arbitrary initial point @xmath51 ( both lying on the physical path ) , and if the constant @xmath52 is taken to be zero ) .
what if , however , some parts of the magnetic field that comprise the magnetic flux @xmath29 actually _ pass through _ some points or a whole region ( interval ) of the path available to the particle ?
in such a case , the above standard argument is not valid ( as @xmath42 can not be written as a grad at any point of the interval where the magnetic field @xmath53 ) .
are there however general results that we can still write for @xmath54 , if the spatial point of observation @xmath49 is again outside the interval with the nonvanishing magnetic field ? or , what if in the previous problems , the magnetic flux ( either remote , or partly passing through the path ) is time - dependent @xmath55 ?
( in that case then , there exists in general an additional electric field @xmath56 induced by faraday s law of induction on points of the path , and the usual gauge transformation argument is once again not valid ) .
returning to another standard ( solvable ) case ( which is actually the dual or the electric analog of the above ) , if along the 1-d physical path the particle experiences only a spatially - uniform ( but generally time - dependent ) classical scalar potential @xmath57 , we can again formally map @xmath58 to a potential - free solution @xmath34 , through a @xmath59 that now solves @xmath60 , and this gives @xmath61 , leading to the electric analog of ( [ standard ] ) , namely @xmath62 with obvious notation .
( notice that , for either of the two mapped systems in this problem , the electric field is zero at all points of the path ) .
what if , however , the scalar potential has also some @xmath49-dependence along the path ( that leads to an electric field ( in a certain interval ) that the particle passes through ) ?
in such a case , the above standard argument is again not valid . are there however general results that we can still write for @xmath63 , if the spatial point of observation @xmath49 is again outside the interval with the nonvanishing electric field ?
we state here directly that this article will provide affirmative answers to questions of the type posed above , by actually giving the corresponding general results in closed analytical forms .
at this point it is also useful to briefly reconsider the earlier mentioned case , namely of a time - dependent @xmath55 that is remote to the 1-d physical path , because in this manner we can immediately provide another motivation for the present work@xmath1 this time - dependent problem is surrounded with a number of important misconceptions in the literature ( the same being true about its electric analog , as we shall see)@xmath1 the formal solution that is usually written down for a @xmath55 is again ( [ standard ] ) , namely the above spatial line integral of @xmath42 , in spite of the fact that @xmath42 is now @xmath27-dependent@xmath7 the problem then is that , because of the first of ( [ basic11 ] ) , @xmath25 must now have a @xmath27-dependence and , from the second of ( [ basic11 ] ) , there must necessarily be scalar potentials involved in the problem ( which have been by force set to zero , in our pre - determined mapping between vector potentials only ) .
having decided to use systems that experience only vector ( and not scalar ) potentials , the correct solution can not be simply a trivial @xmath27-dependent extension of ( [ standard ] ) .
a corresponding error is usually made in the electric dual of the above , namely in cases that involve @xmath48-dependent scalar potentials , where ( [ standard2 ] ) is still erroneously used ( with @xmath64 replaced by @xmath65 ) , giving an @xmath48-dependent @xmath25 , although this would necessarily lead to the involvement of vector potentials ( through the first of ( [ basic11 ] ) and the @xmath48-dependence of @xmath25 ) that have been neglected from the beginning @xmath2 a situation ( and an error ) that appears , in exactly this form , in the description of the so - called electric aharonov - bohm effect@xcite as we shall see .
speaking of errors in the literature , it might here be the perfect place to also point to the reader the most common misleading statement often made in the literature ( and again , for notational simplicity , we restrict our attention to a one - dimensional system , with spatial variable @xmath49 , although the statement is obviously generalizable to ( and often made for systems of ) higher dimensionality by properly using line integrals over arbitrary curves in space)@xmath1 it is usually stated [ e.g. in brown & holland@xcite , see i.e. their eq .
( 57 ) applied for vanishing boost velocity @xmath66 that the general gauge function that connects ( through a phase factor @xmath67 ) the wavefunctions of a quantum system with no potentials ( i.e. with a set of potentials @xmath68 ) to the wavefunctions of a quantum system that moves in vector potential @xmath69 and scalar potential @xmath70 ( i.e. in a set of potentials @xmath71 ) is the obvious combination ( and a natural extension ) of ( [ standard ] ) and ( [ standard2 ] ) , namely @xmath72 which , however , is * incorrect * for @xmath49 and @xmath27 uncorrelated variables@xmath1 it does * not * satisfy the standard system * * * * of gauge transformation equations @xmath73 the reader can easily see why : ( i ) when the @xmath74 operator acts on eq.([wrong ] ) , it gives the correct @xmath75 from the 1st term , but it also gives some annoying additional nonzero quantity from the 2nd term ( that survives because of the @xmath49-dependence of @xmath5 ) ; hence it invalidates the first of the basic system ( [ basicpde ] ) .
( ii ) similarly , when the @xmath76 operator acts on eq.([wrong ] ) , it gives the correct @xmath70 from the 2nd term , but it also gives some annoying additional nonzero quantity from the 1st term ( that survives because of the @xmath27-dependence of @xmath42 ) ; hence it invalidates the second of the basic system ( [ basicpde ] ) .
it is only when @xmath42 is @xmath27-independent , and @xmath5 is spatially - independent , that eq.([wrong ] ) can be correct ( as the above annoying terms do not appear and the basic system is satisfied ) , although it is still not necessarily the most general form for @xmath25 , as we shall see .
[ an alternative form that is also widely thought to be correct is again eq.([wrong ] ) , but with the variables that are not integrated over implicitly assumed to belong to the initial point ( hence a @xmath77 replaces @xmath27 in @xmath42 , and simultaneously an @xmath51 replaces @xmath49 in @xmath5 ) .
however , one can see again that the system ( [ basicpde ] ) is not satisfied ( the above differential operators , when acted on @xmath25 , give @xmath78 and @xmath79 , hence not the values of the potentials at the point of observation @xmath80 as they should ) , this not being an acceptable solution either ] . what is the problem here ? or ,
better put , what is the deeper reason for the above inconsistencies ?
the short answer is the uncritical use of dirac phase factors that come from path - integral treatments .
it is indeed obvious that the form ( [ wrong ] ) that is often used in the literature ( in canonical ( non - path - integral ) formulations where @xmath49 and @xmath27 are * uncorrelated * variables ( and not correlated to produce a path @xmath81 ) ) _ is not generally correct _ , and that is one of the main points that has motivated this work .
we will find _ generalized results _ that actually _ correct _
eq.([wrong ] ) through extra nonlocal terms , and through the proper appearance of @xmath51 and @xmath77 ( as in eq.([lambdastatic1 ] ) and eq.([lambdastatic2 ] ) to be found later in section iii ) , and these are the * exact * ones ( namely the exact @xmath82 that at the end , upon action of @xmath74 and @xmath76 satisfies exactly the basic system ( [ basicpde ] ) , viewed as a system of partial differential equations ( pdes ) ) .
and the formulation that gives these results is generalized later in the article , for @xmath83 ( in the 2-d static case ) and also for @xmath84 ( in the full dynamical 2-d case ) , and leads to the exact ( nontrivial ) forms of the phase function @xmath25 that satisfy ( in all cases ) the system ( [ basicpde ] ) @xmath2 with _ the direct verification _
proof , by going backwards , that these forms are indeed the exact solutions of _ ( [ basicpde ] ) ) _ also being given _ for the reader s convenience__. _ _ [ for the direct and rigorous mathematical derivations see @xcite . ]
this article gives a full exploration of issues related to the above motivating discussion , by pointing to a practical ( and generalized ) use of gauge transformation mapping techniques , that at the end lead to these generalized ( and , at first sight , unexpected ) solutions for the general form of @xmath25 . for cases such as the ones discussed above , or even more involved ones
, there still appears to exist a simple phase mapping ( between two inequivalent systems ) , but the phase connection @xmath25 seems to contain not only integrals of potentials , but also fluxes of the classical field - differences from _ remote _ spacetime regions ( regions , however , that are physically accessible to the particle ) . the above mentioned systems
are the simplest ones where these new results can be applied , but apart from this , the present investigation seems to lead to a number of nontrivial corrections of misleading ( or even incorrect ) reports of the above type in the literature , that are not at all marginal ( and are due to an incorrect use of a path - integral viewpoint in an otherwise canonical framework an error that appears to have been made repeatedly since the original conception of the path - integral formulation , as we shall see ) .
the generalized @xmath25-forms also lead to an honest resolution of earlier paradoxes ( involving relativistic causality ) , and in some cases to a new interpretation of known semiclassical experimental observations , corrections of certain sign - errors in the literature , and nontrivial extensions of earlier semiclassical results to general ( even completely delocalized ) states . [ as a byproduct , we will also show that contrary to what is usually stated in earlier but also recent popular reports the semiclassical phase picked up by classical trajectories ( that are deflected by the lorentz force ) is _ opposite _ ( and not equal ) to the so - called aharonov - bohm phase due to the flux enclosed by the same trajectories@xmath7 we will also provide 2 figures to visually assist the detailed proof of this result as well as to facilitate an elementary physical understanding of this opposite sign relation ] .
most importantly , however , the new formulation seems capable of addressing causal issues in time - dependent single- _ vs _ double - slit experiments , an area that seems to have recently attracted attention@xcite ) .
let us begin with the simplest case of 1-d systems but in the most general dynamic environment , i.e. a single quantum particle of charge @xmath85 , but in the presence of arbitrary ( spatially nonuniform and time - dependent ) vector and scalar potentials .
let us actually consider this particle moving either inside a set of potentials @xmath86 and @xmath87 ( case 1 ) or inside a set of potentials @xmath88 and @xmath89 ( case 2 ) , and try to determine the most general gauge function @xmath90 that takes us from ( maps ) the wavefunctions of the particle in case 1 to those of the same particle in case 2 ( meaning the usual mapping ( [ basic1 ] ) between the wavefunctions of the two systems through the phase factor @xmath91 ) .
[ as already noted , we should keep in mind that for this mapping to be possible we _ must _ assume that at the point @xmath80 of observation ( or measurement of @xmath25 or the wavefunction @xmath92 ) we have equal electric fields ( @xmath93 ) , namely @xmath94 ( so that the @xmath95 s and @xmath5 s in ( [ efield ] ) can indeed satisfy the basic system of equations ( [ basic11 ] ) , or equivalently , of the system of equations ( [ xt - basicsystem ] ) below @xmath2 as can be seen by taking the @xmath96 of the 1st and the @xmath97 of the 2nd of the system ( [ xt - basicsystem ] ) and adding them together)@xmath98 but again , we will _ not _ exclude the possibility of the two systems passing through _ different _ electric fields in other regions of spacetime ( that do _ not _ contain the observation point ) , i.e. for @xmath99 .
in fact , this possibility * will come out naturally * from a careful solution of the basic system ( [ xt - basicsystem])@xmath7 it is for example straightforward for the reader to immediately verify that the results ( [ lambdastatic1 ] ) or ( [ lambdastatic2 ] ) that will be derived below ( and will contain contributions of electric field - differences from remote regions of spacetime ) indeed satisfy the basic input system of equations ( [ xt - basicsystem ] ) , something that will be explicitly verified below ] .
returning to the question on the appropriate @xmath25 that takes us from the set @xmath100 to the set @xmath101 , we note again that , in cases of static vector potentials ( @xmath102 s ) _ and _ spatially uniform scalar potentials ( @xmath57 s ) the form usually given for @xmath25 is the well - known @xmath103 with @xmath104 and @xmath105 ( and it can be viewed as a combination of ( [ standard ] ) and ( [ standard2 ] ) , being immediately applicable to the description of cases of _ combined _ magnetic and electric aharonov - bohm effects ) .
but as already noted , even in the most general case , with @xmath27-dependent @xmath95 s and @xmath49-dependent @xmath5 s ( and with the variables@xmath106 and@xmath107 being * completely uncorrelated * ) , it is often stated in the literature that the appropriate @xmath25 has a form that is a plausible extention of ( [ lambdausual ] ) , namely @xmath108 dx^{\prime}-c{\displaystyle\int\limits_{t_{0}}^{t } } \left [ \phi_{2}(x , t^{\prime})-\phi_{1}(x , t^{\prime})\right ] dt^{\prime } , \label{brownholland}\ ] ] [ with eq . ( 57 ) of ref.@xcite , taken for @xmath109 , being a very good example to point to , since that article does not use a path - integral language , but a canonical formulation with uncorrelated variables ] . and as already pointed out in section ii , this form is certainly _ incorrect _ for uncorrelated variables @xmath49 and @xmath27 ( the reader can easily verify that the system of equations ( [ xt - basicsystem ] ) below is @xmath110 satisfied by ( [ brownholland ] ) @xmath2 see again section ii , especially the paragraph after eq.([basicpde]))@xmath98 we will find in the present work that the correct form consists of two major modifications of ( [ brownholland])@xmath1 ( i ) the first leads to the natural appearance of a _ path _ that continuously connects initial and final points in spacetime , a property that ( [ brownholland ] ) _ does not have _ [ indeed ,
if the integration curves of ( [ brownholland ] ) are drawn in the @xmath80-plane , they do _ not _ form a continuous path from @xmath111 to @xmath80 ] .
( ii ) and the second modification is highly nontrivial@xmath1 it consists of nonlocal contributions of classical electric field - differences from remote regions of spacetime .
we will discuss below the consequences of these terms and we will later show that such nonlocal contributions also appear ( in an extended form ) in more general situations , i.e. they are also present in higher spatial dimensionality ( and they then also involve remote magnetic fields in combination with the electric ones)@xmath7 these lead to modifications of ordinary aharonov - bohm behaviors or have other important consequences , one of them being a natural remedy of causality paradoxes in time - dependent aharonov - bohm experiments .
the form ( [ brownholland ] ) commonly used is of course motivated by the well - known wu & yang@xcite nonintegrable phase factor , that has a phase equal to @xmath113 , a form that appears naturally within the framework of path - integral treatments , or generally in physical situations where narrow wavepackets are implicitly assumed for the quantum particle * * : * * the integrals appearing in ( [ brownholland ] ) are then taken along particle trajectories ( hence spatial and temporal variables _ not _ being uncorrelated , but being connected in a particular manner @xmath81 to produce the path@xmath7 all integrals are therefore basically only time - integrals ) .
but even then , eq.([brownholland ] ) is valid only when these trajectories are always ( in time ) and everywhere ( in space ) inside identical classical fields for the two ( mapped ) systems . here
, however , we will be focusing on what a canonical ( and not a path - integral or other semiclassical ) treatment leads to@xmath7 this will cover the general case of arbitrary wavefunctions that can even be completely spread - out in space , and will also allow the particle to travel through different electric fields for the two systems in remote spacetime regions ( e.g. @xmath114 if @xmath115 etc . ) .
it is therefore clear that finding the appropriate @xmath116 that achieves the above mapping _ in full generality _ will require a careful solution of the system of pdes ( [ basic11 ] ) , applied to only one spatial variable , namely @xmath117 ( with @xmath118 and @xmath119 ) .
this system is underdetermined in the sense that we only have knowledge of @xmath25 at an initial point @xmath111 and with no further boundary conditions ( hence multiplicities of solutions being generally expected , see below ) . by following a careful procedure of integrations@xcite
we finally obtain 2 distinct solutions ( depending on which equation we integrate first)@xmath1 the first solution is @xmath120 with @xmath121 required to be chosen so that the quantity @xmath122 is independent of @xmath49 , and ( from an inverted route of integrations ) the second solution turns out to be @xmath123 with @xmath124 to be chosen in such a way that @xmath125 is independent of@xmath107 . in the above
@xmath126 is the difference of electric fields in the two systems , which can be nonvanishing at regions remote to the observation point @xmath80 ( see examples later below ) .
( note again that at the point of observation @xmath127 , signifying the basic fact that the fields in the two systems are identical at the observation point @xmath80 ) .
the constant last terms in both solutions can be shown to be related to possible multiplicities of @xmath25 ( for a full discussion see @xcite ) and they are zero in simple - connected spacetimes .
also note again that the integrations of potentials in ( [ lambdastatic1 ] ) and ( [ lambdastatic2 ] ) indeed form paths that continuously connect @xmath111 to @xmath80 in the @xmath128-plane ( the red - arrow and green - arrow paths of fig.1(a ) ) , a property that the incorrectly used solution ( [ brownholland ] ) does _ not _ have . by going backwards
one can directly verify that ( [ lambdastatic1 ] ) or ( [ lambdastatic2 ] ) are indeed solutions of the basic system of pdes ( [ xt - basicsystem ] ) , * even for any nonzero * @xmath129 ( in regions @xmath130 ) .
indeed , if we call our first solution ( eq.([lambdastatic1 ] ) ) for simple - connected spacetime @xmath131 , namely @xmath132 with @xmath133 chosen so that @xmath134 is independent of @xmath49 , then we have ( even for @xmath135 for @xmath136)@xmath1 * a ) * @xmath137satisfied trivially@xmath138 ( because the quantity @xmath139 is independent of @xmath49 ) . *
b ) * @xmath140 , ( the last term being trivially zero , @xmath141 ) , and then with the substitution @xmath142 we obtain @xmath143 .
\(i ) we see that the 2nd and 4th terms of the rhs _ cancel each other _ , and \(ii ) the 1st term of the rhs is @xmath144 hence finally @xmath145 we have directly shown therefore that the basic system of pdes ( [ xt - basicsystem ] ) is indeed satisfied by our * generalized * solution @xmath146 * even for any nonzero * @xmath147 ( in regions @xmath99 ) .
( once again note , however , that at the point of observation @xmath127 , indicating the essential fact that the fields in the two systems are equal ( recall that @xmath148 ) at the observation point @xmath80 .
it should be noted that the function @xmath133 owes its existence to the fact that the spacetime point of observation @xmath80 is outside the @xmath56-distribution ( hence the term _ nonlocal , _ used for the effect of the field - difference @xmath56 on the phases ) , and the reader can clearly see this in the striped @xmath56-distributions of the examples that follow later in this section . in a completely analogous way , one can easily see that our alternative solution ( eq.([lambdastatic2 ] ) ) also satisfies the basic system of pdes above .
indeed , if we call our second ( alternative ) solution ( eq.([lambdastatic2 ] ) ) for simple - connected spacetime @xmath149 , namely @xmath150 with @xmath151 chosen so that @xmath152 is independent of @xmath27 , then we have ( even for @xmath135 for @xmath136)@xmath1 * a ) * @xmath153satisfied trivially@xmath138 ( because the quantity @xmath139 is independent of @xmath27 ) .
* b ) * @xmath154 , ( the last term being trivially zero , @xmath155 ) , and then with the substitution @xmath156 we obtain @xmath157 .
\(i ) we see that the 2nd and 4th terms of the rhs _ cancel each other _ , and \(ii ) the 3rd term of the rhs is @xmath158 hence finally @xmath159 once again , all the above are true for any nonzero @xmath129 ( in regions @xmath99 ) for arbitrary analytical dependence of the remote field - difference on its arguments .
figure1.eps let us now note that in ( [ lambdastatic1 ] ) and ( [ lambdastatic2 ] ) the placement of @xmath51 and @xmath77 gives a path - sense to the line integrals in each solution ( each path consisting of 2 perpendicular line segments connecting @xmath111 to @xmath160 with solution ( [ lambdastatic1 ] ) having a clockwise and solution ( [ lambdastatic2 ] ) a counter - clockwise sense , see red and green arrow paths in fig.1)@xmath161this way a natural _ rectangle _ is formed , within which the enclosed electric fluxes in spacetime appear to be crucial ( showing up as nonlocal contributions of the electric field - differences from regions @xmath162 of space and time _ that are remote to the observation point _
@xmath80 ) .
these nonlocal terms in @xmath25 have a direct effect on the wavefunction - phases at @xmath80 .
the actual manner in which this happens is determined by the functions@xmath121 or@xmath124these must be chosen in such a way as to satisfy their respective conditions in ( [ lambdastatic1 ] ) and ( [ lambdastatic2 ] ) . to see how these functions@xmath163 or@xmath164 are actually determined , and what form the above solutions take in nontrivial cases ( and how they give new results , i.e. _ not differing from the usual ones by a mere constant _ )
let us first take examples of striped @xmath56-distributions in spacetime@xmath1 * ( a ) * for the case of the extended _ vertical _ strip ( parallel to the @xmath27-axis ) of fig.1(a ) ( the case of a one - dimensional capacitor that is ( arbitrarily and variably ) charged for all time ) , then , for @xmath49 located outside ( and on the right of ) the capacitor , the quantity@xmath165 in @xmath131 is _
_ already independent of__@xmath106 ( since a displacement of the @xmath80-corner of the rectangle to the right , along the @xmath49-direction , does not change the enclosed electric flux , see fig.1(a))@xmath7 hence in this case the function @xmath133 can be taken as @xmath166 ( up to a constant @xmath167 ) , because then the condition for @xmath133 stated in the solution ( [ lambdastatic1 ] ) ( namely , that the quantity in brackets must be independent of @xmath49 ) is indeed satisfied .
( note again that the above @xmath49-independence of the enclosed electric flux is important for the existence of @xmath133 ) .
so for this setup , the nonlocal term in solution @xmath131 _ survives _ ( the quantity in brackets is nonvanishing ) , but _ _
it is not constant__@xmath1 this enclosed flux depends on @xmath27 ( since the enclosed flux
_ does change _ with a displacement of the @xmath80-corner of the rectangle upwards , along the @xmath27-direction ) .
hence , by looking at the alternative solution @xmath168 the quantity@xmath169 is dependent on@xmath107 , so that @xmath170 must be chosen as @xmath171 ( up to the same constant @xmath167 ) in order to _ cancel _ this @xmath27-dependence , so that its own condition stated in the solution ( [ lambdastatic2 ] ) ( namely , that the quantity in brackets must be independent of @xmath27 ) is indeed satisfied@xmath7 as a result , the quantity in brackets in solution @xmath149 disappears and there is no nonlocal contribution in @xmath149 ( for @xmath172 ) .
( if we had used a @xmath173 , the nonlocal contributions would be differently shared between the two solutions , but without changing the physics when we take the _ difference _ of the two solutions , see below ) . with these choices of @xmath151 and @xmath133
, we already have new results ( compared to the standard ones of the integrals of potentials ) .
i.e. one of the two solutions , namely @xmath131 * is * affected nonlocally by the enclosed flux ( and this flux is * not * constant ) .
spelled out clearly , the two results are : @xmath174 @xmath175 and now it is easy to note that , if we subtract the two solutions @xmath176 and @xmath149 ( and of course assume , as usual , single - valuedness of @xmath25 at the initial point @xmath111 , i.e. @xmath177 ) the result is _ zero _ ( i.e. it so happens that the electric flux determined by the potential - integrals is exactly cancelled by the nonlocal term of electric fields ( i.e. the term that survives in @xmath131 above ) ) , a cancellation effect that is important and that will be generalized in later sections . *
( b ) * in the dual case of an extended _ horizontal _ strip - parallel to the @xmath49-axis ( that corresponds to a nonzero electric field in all space that has however a finite duration @xmath178 , the proper choices ( for observation time instant @xmath179 ) are basically reverse ( i.e. we can now take @xmath180 and @xmath181 ( since the electric flux enclosed in the observation rectangle now depends on @xmath49 , but not on @xmath27 ) , with both choices always up to a common constant ) and once again we can easily see , upon subtraction of the two solutions , a similar cancellation effect . in this case again , the results are also new ( a nonlocal term survives now in @xmath149 ) . again spelled out clearly , these are : @xmath182 @xmath183 their difference also being zero . *
( c ) * and if we want cases that are more involved ( with the nonlocal contributions appearing nontrivially in * both * solutions @xmath131 and @xmath149 and with @xmath133 and @xmath151 not being immediately visible ) we must again consider different shapes of @xmath56-distribution .
one such case ( a triangular @xmath56-distribution ) is shown in fig.1(b ) ( for the corresponding magnetic case to be discussed in the next section , which however is completely analogous ) ; in this case the enclosed flux depends on _ both _ @xmath49 and @xmath27 ( but can be shown to be separable , so that the functions @xmath133 and @xmath151 still exist and can easily be found in closed analytical form , see next section ) . as for the last constant terms @xmath184 and @xmath185 ( what we will call multiplicities ) , these are only present ( nonvanishing )
when @xmath25 is expected to be multivalued , i.e. in cases of motion in multiple - connected spacetimes , and are then related to the fluxes in the inaccessible regions : in the electric aharonov - bohm setup , the prototype of multiple - connectivity in spacetime@xcite , it turns out@xcite that @xmath186 enclosed ( and here inaccessible ) electric flux , and if these values are substituted in ( [ lambdastatic1 ] ) and ( [ lambdastatic2 ] ) they cancel out the new nonlocal terms and lead to the usual electric aharonov - bohm result ( of mere integrals over potentials ) . as for other , more esoteric properties of the new solutions in simple - connected spacetimes
, it can be rigorously shown@xcite that solutions ( [ lambdastatic1 ] ) and ( [ lambdastatic2 ] ) are actually _ equal _ , because @xmath133 turns out to be equal to the @xmath27-independent bracket of ( [ lambdastatic2 ] ) , and @xmath170 turns out to be equal to the @xmath49-independent bracket of ( [ lambdastatic1 ] ) , the nonlocal terms having therefore the tendency to exactly cancel the aharonov - bohm terms ( this being true _ for arbitrary shapes and analytical form of _
@xmath187 ) .
after having discussed fully the simple @xmath80-case , let us for completeness give the analogous ( euclidian - rotated in 4-d spacetime ) derivation for @xmath188-variables and briefly discuss the properties of the simpler static solutions in spatial two - dimensionality .
we will simply need to apply the same methodology ( of solution of a system of pdes ) to such static spatially two - dimensional cases ( so that now different ( remote ) magnetic fields for the two systems , perpendicular to the 2-d space , will arise ) .
for such cases we need to solve the system of pdes that is now of the form @xmath189 by following then a similar procedure of integrations@xcite we obtain the following general solution@xmath190 @xmath191 a result that applies to cases where the particle goes through _ different _ perpendicular magnetic fields @xmath192 and @xmath193 _ in spatial regions that are remote to ( i.e. do not contain ) the observation point _
@xmath188 ( and in the above @xmath194 ) .
the reader should note that the first 3 terms of ( [ lambda(x , y)new ] ) are the ( total ) dirac phase along two perpendicular segments that continuously connect the initial point @xmath195 to the point of observation @xmath188 , _ in a clockwise sense _
( see for example the red - arrow paths in fig.1(b ) ) .
but apart from this dirac phase , we also have nonlocal contributions from @xmath196 and its flux within the observation rectangle ( see i.e. the rectangle being formed by the red- and green - arrow paths in fig.1(b ) ) .
below we will directly verify that ( [ lambda(x , y)new ] ) is indeed a solution of ( [ usualgrad ] ) ( even for @xmath197 for @xmath198 ) .
alternatively , by following the reverse route of integrations , we finally obtain the following alternative general solution @xmath199 @xmath200 and again the reader should note that , apart from the first 3 terms ( the ( total ) dirac phase along the two other ( alternative ) perpendicular segments ( connecting @xmath195 to @xmath188 ) , now _ in a counterclockwise sense _ ( the green - arrow paths in fig.1(b ) ) ) , we also have nonlocal contributions from the flux of @xmath196 that is enclosed within the same observation rectangle ( that is naturally defined by the four segments of the two solutions ( fig.1(b ) ) ) . in all the above , @xmath201 and @xmath202 are the cartesian components of @xmath203 , and , as already mentioned , @xmath196 is the difference between ( perpendicular ) magnetic fields that the two systems may experience in regions that _ do not contain _ the observation point @xmath188 ( i.e. @xmath204 @xmath205 , and , although at the point of observation @xmath188 we have @xmath206 ( already emphasized in the introductory sections ) , this @xmath207 can be nonzero for @xmath208 ) . it should be noted that it is because of @xmath206 that the functions @xmath133 and @xmath209 of ( [ lambda(x , y)new ] ) and ( [ lambda(x , y)4 ] ) can be found , and the new solutions therefore exist ( and are nontrivial ) . for the impatient reader ,
simple physical examples with the associated analytical forms of @xmath133 and @xmath209 ( derived in detail ) are given later in this section .
one can again show that the 2 solutions are equal for simple - connected space ( when the last constant terms @xmath210 and @xmath211 are vanishing ) , and for multiple - connectivity the values of the multiplicities @xmath210 and @xmath211 cancel out the nonlocalities and reduce the above to the usual result of mere @xmath95-integrals along the 2 paths ( i.e. two simple dirac phases ) .
a direct backwards verification that ( [ lambda(x , y)new ] ) and ( [ lambda(x , y)4 ] ) do indeed satisfy the basic system ( [ usualgrad ] ) * ( even for cases with * @xmath212 in remote regions of space * * ) * * can be made along similar lines to the ones of the last section@xmath1 for simple - connected space , let us call our solution ( [ lambda(x , y)new ] ) @xmath213 , namely @xmath214 with @xmath133 chosen so that @xmath215is independent of @xmath216 we then have ( even for @xmath197 for @xmath217)@xmath1 * * a ) * * @xmath218satisfied trivially@xmath138 ( because the quantity @xmath219 is independent of @xmath49 ) .
* * b ) * * @xmath220 ( the last term being trivially zero , @xmath221 ) , and then with the substitution @xmath222 we obtain @xmath223 \(i ) we see that the 2nd and 4th terms of the right - hand - side ( rhs ) _ cancel each other _ , and \(ii ) the 1st term of the rhs is @xmath224 hence finally @xmath225 we have directly shown therefore ( by going backwards ) that the basic system of pdes ( [ usualgrad ] ) is indeed satisfied by our generalized solution @xmath226 even for any nonzero @xmath207 ( in regions @xmath227 recall that always @xmath206 ) . to fully appreciate the above simple proof , the reader is again urged to look at the cases of striped @xmath196-distributions later below , the point of observation @xmath188 always lying outside the strips , so that the above function @xmath133 can easily be determined , and the new solutions really _ exist _ - and they are nontrivial . in a completely analogous way
, one can easily see that our alternative solution ( eq.([lambda(x , y)4 ] ) ) also satisfies the basic system of pdes above . indeed ,
if we call this second static solution @xmath228 , namely @xmath229 with @xmath209 chosen so that @xmath230 is independent of@xmath231 , then we have ( even for @xmath197 for @xmath208)@xmath1 * * a ) * * @xmath232satisfied trivially@xmath138 ( because the quantity @xmath219 is independent of @xmath231 ) . * * b ) * * @xmath233 ( the last term being trivially zero , @xmath234 ) , and then with the substitution @xmath235 we obtain @xmath236 \(i ) we see that the last two terms of the rhs _ cancel each other _ , and \(ii ) the 2nd term of the rhs is @xmath237 hence finally @xmath238 once again , all the above are true for any nonzero @xmath239 ( in regions @xmath198 ) for arbitrary analytical dependence of the remote field - difference on its arguments . and for a clearer understanding of this proof let us now turn to the striped examples promised earlier . in order to see again how the above solutions appear in nontrivial cases ( and how they give completely new results , i.e. _ not differing from the usual ones _ ( _ i.e . from the dirac phase _ ) _ by a mere constant _ ) let us first take examples of striped @xmath196-distributions in space@xmath1 * ( a ) * for the case of an extended _ vertical _ strip - parallel to the @xmath231-axis , such as in fig.1(a ) ( imagine @xmath27 replaced by @xmath231 ) ( i.e. for the case that the particle has actually passed through nonzero @xmath196 , hence through _ different _ magnetic fields in the two ( mapped ) systems ) , then , for @xmath49 located outside ( and on the right of ) the strip , the quantity@xmath240 in @xmath213 _ _ is already independent of__@xmath106 ( since a displacement of the @xmath188-corner of the rectangle to the right , along the @xmath49-direction , does not change the enclosed magnetic flux @xmath2 see fig .
1(a ) for the analogous @xmath80-case discussed earlier)@xmath98 indeed , in this case the above quantity ( the enclosed flux within the observation rectangle ) does not depend on the @xmath49-position of the observation point , but on the positioning of the boundaries of the @xmath196-distribution in the @xmath49-direction ( better , on the constant width of the strip ) @xmath2 as the @xmath49-integral does not give any further contribution when the dummy variable @xmath241 goes out of the strip .
in fact , in this case the enclosed flux depends on @xmath231 as we discuss below ( but , again , not on @xmath49 ) .
hence , for this case , the function @xmath133 can be easily determined : it can be taken as @xmath166 ( up to a constant @xmath167 ) , because then the condition for @xmath133 stated in solution ( [ lambda(x , y)new ] ) ( namely , that the quantity in brackets must be independent of @xmath49 ) is indeed satisfied .
we see therefore above that for this setup , the nonlocal term in solution @xmath213 _ survives _ ( the quantity in brackets is nonvanishing ) , but _
_ it is not constant__@xmath1 as already noted , this enclosed flux depends on @xmath231 ( since the enclosed flux _
does change _ with a displacement of the @xmath188-corner of the rectangle upwards , along the @xmath231-direction , as the @xmath231-integral _ is _ affected by the positioning of @xmath231 @xmath2 the higher the positioning of the observation point the more flux is enclosed inside the observation rectangle ) .
hence , by looking at the alternative solution @xmath242 the quantity@xmath243 is _ _ dependent on__@xmath244 , so that @xmath209 must be chosen as @xmath245 ( up to the same constant @xmath167 ) in order to _ cancel _ this @xmath231-dependence , so that its own condition stated in solution ( [ lambda(x , y)4 ] ) ( namely , that the quantity in brackets must be independent of @xmath231 ) is indeed satisfied@xmath7 as a result , the quantity in brackets in solution @xmath228 disappears and there is no nonlocal contribution in @xmath228 ( for @xmath172 ) .
( if we had used a @xmath173 , the nonlocal contributions would be shared between the two solutions in a different manner , but without changing the physics when we take the _ difference _ of the two solutions ( see below ) ) .
[ the crucial point in the above for the existence of @xmath133 and @xmath209 is , once again , the fact that @xmath246 at @xmath188 , combined with the sharp boundaries of the nonvanishing @xmath196-region ] . with these choices of @xmath209 and @xmath133
, we already have new results ( compared to the standard ones of the integrals of potentials ) .
i.e. one of the two solutions , namely @xmath213 * is * affected nonlocally by the enclosed flux ( and this flux is * not * constant ) .
spelled out clearly , the two results are : @xmath247 @xmath248 and now it is easy to note that , if we subtract the two solutions @xmath249 and @xmath228 , the result is _ zero _ ( because the line integrals of the vector potential @xmath4 in the two solutions are in opposite senses in the @xmath188 plane , hence their difference leads to a _ closed _ line integral of @xmath4 , which is in turn equal to the enclosed magnetic flux , and this flux always happens to be of opposite sign from that of the enclosed flux that explicitly appears as a nonlocal contribution of the @xmath196-fields ( i.e. the term that survives in @xmath213 above)@xmath98 hence , the two solutions are _
equal_. [ we of course everywhere assumed , as usual , single - valuedness of @xmath25 at the initial point @xmath195 , i.e. @xmath250 matters of multivaluedness of @xmath25 at the observation point @xmath188 will be addressed later below ] .
it is interesting that , formally speaking , the above _ equality _ of the two solutions is due to the fact that the @xmath49-independent quantity in brackets of the 1st solution ( [ lambda(x , y)new ] ) is equal to the function @xmath209 of the 2nd solution ( [ lambda(x , y)4 ] ) , and the @xmath231-independent quantity in brackets of the 2nd solution ( [ lambda(x , y)4 ] ) is equal to the function @xmath133 of the first solution ( [ lambda(x , y)new ] ) .
this turns out to be a general behavioral pattern of the two solutions in simple - connected space , that is valid for any shape ( and any analytical form ) of @xmath196-distribution .
but most importantly , it should be noted that this vanishing of @xmath251 is a cancellation effect that is emphasized further later below and discussed in completely _ physical terms_. * ( b ) * in the dual case of an extended _ horizontal _ strip - parallel to the @xmath49-axis , the proper choices ( for @xmath231 above the strip ) are basically reverse ( i.e. we can now take @xmath252 and @xmath253 @xmath254(since the flux enclosed in the rectangle now depends on @xmath49 , but not on @xmath231 ) , with both choices always up to a common constant ) and once again we can easily see , upon subtraction of the two solutions , a similar cancellation effect . in this case
as well , the results are again new ( a nonlocal term survives now in @xmath228 ) . again spelled out clearly , these are : @xmath255 @xmath256 ( their difference also being zero ) .
again here the crucial point is that , because the @xmath196-configuration does _ not contain the point _
@xmath188 , a displacement of this observation point upwards does _ not _ change the flux inside the observation rectangle@xmath7 this makes the new solutions ( i.e the functions @xmath133 and @xmath209 ) exist . *
( c ) * and if we want cases that are more involved ( i.e. with the nonlocal contributions appearing nontrivially in * both * solutions @xmath213 and @xmath228 and with @xmath133 and @xmath209 not being immediately visible ) , we must consider different shapes of @xmath196-distributions .
one such case is a triangular one that is shown in fig.1(b ) ( for simplicity an equilateral triangle , and with the initial point @xmath257 ) and with the point of observation @xmath188 being fairly close to the triangle s right side as in the figure .
note that for such a configuration , the part of the magnetic flux that is inside the observation rectangle ( defined by the right upper corner @xmath188 ) depends on * both * @xmath49 * and * @xmath231 .
it turns out , however , that this @xmath258 and @xmath259dependent enclosed flux can be written as a sum of separate @xmath49- and @xmath231-contributions , so that appropriate @xmath133 and @xmath209 can still be found ( each one of them must be chosen so that it only cancels the corresponding variable s dependence of the enclosed flux ) . for a homogeneous @xmath196
it is a rather straightforward exercise to determine this enclosed part , i.e. the common area between the observation rectangle and the equilateral triangle , and from this we can find the appropriate @xmath133 that will cancel the @xmath49-dependence , and the appropriate @xmath209 that will cancel the @xmath231-dependence .
these appropriate choices turn out to be @xmath260 + c \label{triangular1}\ ] ] and @xmath261 + c \label{triangular2}\ ] ] with @xmath262 being the side of the equilateral triangle . we should emphasize that expressions ( [ triangular1 ] ) and ( [ triangular2 ] ) , if combined with ( [ lambda(x , y)new ] ) or ( [ lambda(x , y)4 ] ) , give the nontrivial nonlocal contributions of the field - difference @xmath196 of the remote magnetic fields on @xmath25 of each solution ( hence on the phase of the wavefunction of each wavepacket travelling along each path ) at the observation point @xmath188 , that always lies outside the @xmath196-triangle .
( we mention again that in the case of completely spread - out states , the equality of the two solutions at the observation point essentially demonstrates the uniqueness ( single - valuedness ) of the phase in simple - connected space ) .
further physical discussion of the above cancellations , and a semiclassical interpretation , is given later below .
finally , in more difficult geometries , i.e. when the shape of the @xmath196-distribution is such that the enclosed flux does _ not _ * * decouple in a sum of separate @xmath49- and @xmath231-contributions , such as cases of circularly shaped @xmath196-distributions , it is advantageous to solve the system ( [ usualgrad ] ) directly in non - cartesian ( i.e. polar ) coordinates@xcite .
a general comment that can be made for general shapes is that , depending on the geometry of shape of the @xmath196-distribution , an appropriate change of variables ( to a new coordinate system ) may first be needed , so that generalized solutions of the system ( [ usualgrad ] ) can be found ( namely , so that the enclosed flux inside the _ transformed _ observation rectangle ( i.e. a slice of an annular section in the case of polar coordinates ) can be written as a sum of separate ( transformed ) variables ) , and then the same methodology ( as in the above cartesian cases ) can be followed .
as we saw in the above examples , in case of a striped - distribution of the magnetic field difference @xmath196 , the functions @xmath133 and @xmath209 in ( [ lambda(x , y)new ] ) and ( [ lambda(x , y)4 ] ) have to be chosen in ways that are compatible with their corresponding constraints ( stated after ( [ lambda(x , y)new ] ) and ( [ lambda(x , y)4 ] ) ) and are completely analogous to the above discussed @xmath80-cases@xmath98 by then taking the _ difference _ of ( [ lambda(x , y)new ] ) and ( [ lambda(x , y)4 ] ) ) we obtain that the aharonov - bohm phase ( the one originating from the _ closed _ line integral of @xmath95 s ) is exactly cancelled by the additional nonlocal term of the magnetic fields ( that the particle passed through ) .
this is quite reminiscent of the cancellation of phases observed in the early experiments of werner & brill@xcite for particles passing through a magnetic field ( a cancellation between the aharonov - bohm phase and the semiclassical phase picked up by the trajectories ) , and our method seems to provide a natural explanation@xmath1 as our results are general ( and for delocalized states in simple - connected space they basically demonstrate the uniqueness of @xmath25 ) , they are also valid and applicable to states that describe wavepackets in classical motion , as _ was _ the case of the werner & brill experiments .
( a similar cancellation of an electric aharonov - bohm phase , that has never been noted in the literature , also occurs for particles passing through a static electric field , and this we will independently prove below , again for semiclassical states ) . we conclude that , for static cases , and when particles pass through nonvanishing classical fields , the new nonlocal terms reported in the present work lead quite generally to a cancellation of aharonov - bohm phases that had earlier been only sketchily noticed and only at the semiclassical ( magnetic ) level .
figure2.eps what is however more important to point out here is that the above cancellations can be understood as a compatibility between the aharonov - bohm fringe - displacement and the trajectory - deflection due to the lorentz force ( i.e. the semiclassical phase picked up due to the optical path difference of the two deflected trajectories _ exactly cancels _ ( is _ opposite in sign _ from ) the aharonov - bohm phase picked up by the same trajectories due to the flux that they enclose ) .
this opposite sign seems to have been rather unnoticed@xmath1 in feynman s fig.15 - 8 @xcite , or in felsager s
fig.2.16 @xcite , the classical trajectories are deflected after passing through a strip of a magnetic field placed on the right of a double - slit apparatus .
both authors determine the semiclassical phase picked up by the deflected trajectories and find it consistent with the aharonov - bohm phase .
one can see on closer inspection , however , that the two phases actually _ have opposite signs _
( see our own fig.2 and the discussion that follows below , where this is proved in detail ) .
similarly , in the very recent review of batelaan & tonomura@xcite , their fig.2 shows wavefronts associated to the deflected classical trajectories where it is stated that the phase shift calculated in terms of the lorentz force is the same as that predicted by the aharonov - bohm effect in terms of the vector potential .
once more , however , it turns out that the sign of the classical phase - difference is really opposite to the sign of the aharonov - bohm phase ( see proof below ) .
the phases are not equal as stated by the authors . and
it turns out that even electric analogs of the above cases also demonstrate this opposite - sign relationship ( see proof further below ) .
all the above examples can be viewed as a manifestation of the cancellations that have been found in the present work for _ general _ ( even completely spread - out ) quantum states ( but in those examples they are just special cases for narrow wavepacket - states in classical motion ) .
let us give a brief elementary proof of the above claimed opposite sign - relationships@xmath1 indeed , in our fig.2 , the aharonov - bohm phase due to the flux enclosed between the two classical trajectories ( of a particle of charge @xmath10 ) is @xmath263 with @xmath264 the flux quantum , and @xmath265 the enclosed flux between the two trajectories ( for small trajectory - deflections ) , with the deflection originating from the presence of the magnetic strip @xmath266 and the associated lorentz forces . on the other hand ,
the semiclassical phase difference between the same 2 classical trajectories is @xmath267 , with @xmath268 being the de broglie wavelength ( and @xmath269 being the speed of the particle , taken almost constant ( as usually done ) due to the small deflections ) , and with @xmath270 being @xmath271 ( @xmath272 being the ( displaced ) position of the central fringe on the screen ) .
we have therefore @xmath273 now , the lorentz force ( exerted only during the passage through the thin magnetic strip , hence only during a time interval @xmath274 ) has a component parallel to the screen ( let us call it @xmath49-component ) that is given by @xmath275 which shows that there is a change of kinematic momentum ( parallel to the screen ) equal to @xmath276 or , equivalently , a change of parallel speed @xmath277 which is the speed of the central fringe s motion ( i.e. its displacement over time along the screen ) .
although this has been caused by the presence of the thin deflecting magnetic strip , this displacement is occuring uniformly during a time interval @xmath278 and this time interval must satisfy @xmath279 ( as , for small displacements , the wavepackets travel most of the time in uniform motion , i.e. @xmath280 ) .
we therefore have that the central fringe displacement must be @xmath281 and noting that @xmath282 , we finally have @xmath283 by susbstituting ( [ xc ] ) into ( [ semiphase ] ) , the lengths @xmath284 and @xmath285 cancel out , and we finally have @xmath286 which with @xmath287 the flux quantum , and @xmath288 the enclosed flux ( always for small trajectory - deflections ) gives ( through comparison with ( [ abphase ] ) ) our final proof that @xmath289 [ it should be noted that such an opposite sign relation actually tells us that the _ total _ ( semiclassical + ab ) _ _ phase _ difference _ is zero at the new position of the central ( bright ) fringe ( after it has been displaced due to the trajectories deflection ) , and in this sense the above minus sign should be rather expected . ]
figure3.eps the electric analog of the above exercise is also outlined below , now with a homogeneous electric field ( pointing downwards everywhere in space , but switched on for only a finite duration @xmath290 ) on the right of a double - slit apparatus ( see our fig.3)@xmath1 in this case the electric lorentz force @xmath291 is exerted on the trajectories only during the small time interval @xmath292 which we take to be much shorter ( @xmath293 ) than the time of travel @xmath294 ( we now have a thin electric strip in _ time _ rather than the thin magnetic strip in space that we had earlier ) .
the electric type of aharonov - bohm phase is now @xmath295 with @xmath296 being the electric potential difference between the two trajectories , hence @xmath297 ( again for small trajectory - deflections ) . on the other hand , the semiclassical phase difference between the two trajectories is again given by ( [ semiphase ] ) , but the position @xmath272 of the central fringe
must now be determined by the electric field force @xmath298 the change of kinematic momentum ( always parallel to the screen ) is now @xmath299 , hence the analog of ( [ dvx ] ) is now @xmath300 which if combined with ( [ dvx2 ] ) ( that is obviously valid in this case as well , again for small deflections , due to the @xmath301 ) , and always with @xmath294 , gives that the central fringe displacement must be @xmath302 , and using again @xmath303 , we finally have the following analog of ( [ xc ] ) @xmath304 by substituting ( [ xcnew ] ) into ( [ semiphase ] ) , the lengths @xmath284 and @xmath285 again cancel out , and we finally have @xmath305 , which with @xmath287 the flux quantum , and through comparison with ( [ abele ] ) leads once again to our final proof that @xmath306 we note therefore that even in the electric case , the semiclassical phase difference ( between two trajectories ) picked up due to the lorentz force ( exerted on them ) is once again opposite to the electric aharonov - bohm phase phase picked up by the same trajectories ( due to the electric flux that they enclose )
. we should point out once again , however , that although the above elementary considerations apply to semiclassical motion of narrow wavepackets , in this paper we have given _ a more general understanding of the above opposite sign - relationships _ that applies to general ( even completely delocalized ) states , and that originates from what could be called generalized werner & brill cancellations ( the cancellations that come out of our new solutions , in cases more general than those of werner & brill , including both magnetic and electric fields , both static and @xmath27-dependent situations , and both semiclassical and spread - out quantum states ) . in a slightly different vein , the cancellations that we found above
give an explanation of why certain classical arguments ( invoking the past @xmath27-dependent history of an experimental setup ) seem to be successful in giving at the end an explanation of aharonov - bohm effects ( namely a phase consistent with that of a static aharonov - bohm configuration ) .
however , there is again an opposite sign that seems to have been largely unnoticed in such arguments as well ( i.e. in silverman@xcite , where in his eq.(1.34 ) there should be an extra minus sign ) .
our above observation essentially describes the fact that , _ if we had actually used _ a @xmath27-dependent magnetic flux ( with its final value being the actual value of our static flux )
, then the induced electric field ( viewed now as a nonlocal term of the present work ) would have cancelled the static aharonov - bohm phase .
of course now , this @xmath27-dependent experimental set up has not been used ( the flux is static ) and we obtain the usual magnetic aharonov - bohm phase , but the above argument ( of a potential experiment that _ could have been carried out _ ) takes the mystery away of why such history - based arguments generally work @xmath2 although _ they have to be corrected with a sign_. the above also gives a rather natural account of the dynamical nonlocality character@xcite attributed to the various aharonov - bohm phenomena ( magnetic , electric or combined ) , although @xmath2 in the present work @xmath2 this dynamical quantum nonlocality seems to simultaneously respect causality , as will be seen in the next section .
this is a rather pleasing characteristic of this theory that , as far as we are aware , has no parallel in other formulations .
finally , let us look at the most general spatially - two - dimensional and time - dependent case .
this combines effects of ( perpendicular ) magnetic fields ( which , if present only in physically - inaccessible regions , can have aharonov - bohm consequences ) with the temporal nonlocalities of electric fields ( parallel to the plane ) found in previous sections . by working again in cartesian spatial coordinates , we now have to deal with the full system of pdes @xmath307 this exercise is considerably longer than the previous ones but important to solve , in order to see in what manner the solutions of this system are able to _ combine _ the spatial and temporal nonlocal effects found above .
there are now 3!=6 alternative integration routes to follow for solving this system ( and , in addition to this , the results in intermediate steps tend to proliferate ) .
the corresponding ( rather long ) procedure for solving the system ( [ fullsystem ] ) is described in detail in @xcite , and 2 out of the 12 solutions that can be derived turn out to be the most crucial for the discussion that will follow .
first , by following steps similar to the ones of section iv , the following temporal generalization of ( [ lambda(x , y)4 ] ) is obtained @xmath308 @xmath309 @xmath310 and from this point on , the third equation of the system ( [ fullsystem ] ) is getting involved to determine the nontrivial effect of scalar potentials on @xmath311 indeed , by combining it with ( [ lambda(x , y , t)213 ] ) there results a wealth of patterns , one of them leading finally to our first solution , namely @xmath312 @xmath313 with the functions @xmath314 and @xmath315 to be chosen in such a way as to satisfy the following 3 independent conditions@xmath1 @xmath316 which is of course a special case of the condition on @xmath317 above ( see after ( [ lambda(x , y , t)213 ] ) ) applied at @xmath318 , and the other 2 turn out to be of the form @xmath319 @xmath320 ( it is probably important to note that for the above results the faraday s law is crucial@xcite ) . as for the constant quantity @xmath321 appearing in ( [ lambdafull1 ] ) , it again describes possible effects of multiple - connectivity at the instant @xmath77 ( which are absent for simple - connected spacetimes , but will be crucial in the discussion of the van kampen thought - experiment to be discussed later below ) .
( [ lambdafull1 ] ) is our first solution .
it is now crucial to note that an alternative form of solution ( with the functions @xmath322 and @xmath323 satisfying the _ same _ conditions as above ) can be derived , and it turns out to be @xmath324 @xmath325 in this alternative solution we note that , in comparison with ( [ lambdafull1 ] ) , the line - integrals of @xmath326 have changed to the _ other _ alternative path ( note the difference in the placement of the coordinates of the initial point @xmath195 in the arguments of @xmath327 and @xmath328 compared to ( [ lambdafull1 ] ) ) and they happen to have the same sense as the @xmath4-integrals , while simultaneously the magnetic flux difference shows up with its value at the initial time @xmath77 rather than at @xmath27 .
this alternative form will be shown to be useful in cases where we want to directly compare physical situations in the present ( at time @xmath27 ) and in the past ( at time @xmath77 ) , and the above noted change of sense of @xmath329-integrals ( compared to ( [ lambdafull1 ] ) ) will be crucial in the discussion that follows in the next section .
( it is also important here to note that , in the form ( [ lambdafull2 ] ) , the electric fields have already incorporated the effect of radiated @xmath196-fields in space ( through the maxwell s equations , see @xcite ) , and this is why at the end only the @xmath196 at @xmath77 appears explicitly ) .
once again the reader can directly verify that ( [ lambdafull1 ] ) or ( [ lambdafull2 ] ) indeed satisfy the basic input system ( [ fullsystem ] ) .
( this verification is considerably more tedious than the earlier ones but rather straightforward ) . but a last mathematical step remains@xmath1 in order to discuss the van kampen case , namely an enclosed ( and physically inaccessible ) magnetic flux ( which however is _ time - dependent _ ) _ , _ it is important to have the analogous forms through a reverse route of integrations ( see @xcite ) , where at the end we will have the reverse path of @xmath4-integrals ( so that by taking the _ difference _ of the resulting solution and the above solution ( [ lambdafull1 ] ) ( or ( [ lambdafull2 ] ) ) will lead to the _ closed _ line integral of @xmath4 , * * which will then be immediately related to van kampen s magnetic flux ( at the instant @xmath27 ) ) . by following then the reverse route , and by applying a similar strategy at every intermediate step ,
we finally obtain the following solution ( the spatially dual of ( [ lambdafull1 ] ) ) , namely @xmath330 @xmath331 with the functions @xmath332 and @xmath315 to be chosen in such a way as to satisfy the following 3 independent conditions@xmath1 @xmath333 @xmath334 @xmath335 where again for the above results the faraday s law is crucial .
the corresponding analog of the alternative form ( [ lambdafull2 ] ) ( where @xmath196 appears at @xmath77 ) is more important for the discussion that follows and turns out to be @xmath336 @xmath337 with @xmath332 and @xmath338 following the same 3 conditions above .
the constant term @xmath339 again describes possible multiplicities at the instant @xmath340 it is absent for simple - connected spacetimes , but will be crucial in the discussion of the van kampen thought - experiment . in ( [ lambdafull4 ] ) ( and in ( [ lambdafin ] ) ) , note the alternative paths ( compared to solution ( [ lambdafull1 ] ) ( and ( [ lambdafull2 ] ) ) ) of line integrals of @xmath4 s ( or of @xmath329 s ) . but
the most crucial element for what follows is the need to _ exclusively _ use the forms ( [ lambdafull2 ] ) and ( [ lambdafin ] ) ( where @xmath196 only appears at @xmath77 ) , and the fact that , within each solution , the sense of @xmath4-integrals is the _ same _ as the sense of the @xmath329-integrals .
( this is _ not _ true in the other solutions where @xmath341 appears , as the reader can directly see ) .
these facts will be crucial to the discussion that follows , which briefly addresses the so called van kampen paradox . in @xcite van kampen considered a magnetic aharonov - bohm setup , but with an inaccessible magnetic flux that is @xmath27-dependent@xmath1 he envisaged turning on the flux very late , or equivalently , observing the interference of the two wavepackets on a distant screen very early , earlier than the time it takes light to travel the distance to the screen ( i.e. @xmath342 ) , hence using the ( instantaneous nature of the ) aharonov - bohm phase to transmit information ( on the presence of a confined flux somewhere in space ) _
superluminally_. indeed , the aharonov - bohm phase at any @xmath27 is determined by differences of @xmath343 with @xmath344 ( basically a special case of ( [ wrong ] ) )
. however , if we use , instead , our results ( [ lambdafull2 ] ) and ( [ lambdafin ] ) above ( that contain the additional nonlocal terms ) , it turns out that , for a spatially - confined magnetic flux @xmath55 , the functions @xmath345 @xmath346 and @xmath323 in the above solutions can then all be taken zero@xmath1 * ( i ) * their conditions are all satisfied for a flux @xmath55 that is not spatially - extended ( hence , from the 2 conditions on @xmath322 and @xmath346 ( eq .
( [ gcondition ] ) and ( [ gcondition3 ] ) ) we obtain @xmath347 since the integrals in brackets are all independent of @xmath49 and @xmath231 ) , and * ( ii ) * for @xmath348 @xmath349 , the integrals of @xmath327 and @xmath328 in the corresponding conditions ( eq.([f(x , y)condition1 ] ) and ( [ f(x , y)condition2 ] ) ) are already independent of _ both _ @xmath49 and @xmath231 ( since @xmath350 for all @xmath351 , with @xmath188 the observation point [ the essential point being that at instant @xmath27 , the @xmath329-field has not yet reached the spatial point @xmath188 of the screen , and therefore all integrations of @xmath327 and @xmath328 with respect to @xmath241 and @xmath352 will be contributing only up to a light - cone ( see fig.4 ) and they will therefore give results that are _ independent of the integration upper limits _ @xmath49 _ and _
@xmath231 @xmath2 basically a generalization of the striped cases that we saw earlier but now to the case of 3 spatio - temporal variables ( with now the spatial point @xmath188 being outside the light - cone defined by @xmath27 ( see fig.4 ; in this figure the initial spatial point @xmath195 , taken for simplicity at @xmath353 , has been supposed to be in the area of the inaccessible flux @xmath55 , so that , for @xmath354 , we have indeed that @xmath355 , as written on the figure))]@xmath7 we therefore rigorously obtain @xmath356 ) .
moreover , the aharonov - bohm multiplicities ( at @xmath77 ) lead to cancellation of the @xmath196-terms ( always at @xmath77 ) , with the final result ( after subtraction of the 2 solutions ) being @xmath357
@xmath358 this can equivalently be written as@xmath359 which , with @xmath360 the instantaneous enclosed magnetic flux and with the help of faraday s law @xmath361 gives @xmath362 although @xmath363 is generally @xmath27-dependent , we obtain the intuitive ( causal ) result that , for @xmath364 ( i.e. if the physical information has not yet reached the screen ) , the phase - difference turns out to be @xmath27-independent , and leads to the magnetic aharonov - bohm phase that we _ would _ observe at @xmath77 . _ the new nonlocal terms have conspired in such a way as to exactly cancel the causality - violating aharonov - bohm phase _
( that would be proportional to the instantaneous @xmath55 ) .
this gives an honest resolution of the van kampen paradox within a canonical formulation , without using any vague electric aharonov - bohm argument ( as there is no multiple - connectivity in the @xmath80-plane@xcite ) .
an additional physical element is that , for the above cancellation , it is not only the @xmath56-fields but also the @xmath27-propagation of the @xmath196-fields ( the full radiation field ) that plays a role@xcite .
figure4.eps use of other 10 solutions that rigorously come out from the basic system of pdes can also address bound - state analogs ( i.e. in @xmath27-dependent magnetic flux - driven 1-d nanorings , such as in @xcite ) or even electric analogs of the above van kampen case@xmath1 in peshkin s review@xcite on the electric aharonov - bohm effect,@xmath0the author correctly states one can not wait for the electron to pass and only later switch on the field to cause a physical effect .
although peshkin uses his eq.(b.5 ) and ( b.6 ) ( based on the incorrect ( [ wrong ] ) ) , he carefully states that it is not the full solution ; actually , if we view it as an _ ansatz _ , then it is understandable why he needs to enforce a _ condition _ ( his eq.(b.8 ) , and later ( b.9 ) ) on the electric field outside the cages ( in order for certain ( annoying ) terms ( resulting from a minimal substitution due to the incorrect ansatz ) to vanish and for ( b.5 ) to be a solution ) .
but then he notes that the extra condition can not always be satisfied ( hence ( b.5 ) is not really the solution for all times ) , drawing from this the above qualitatively correct conclusion on causality .
as it turns out , our treatment gives exactly what peshkin describes in words ( with the total radiation field outside the cages being once again crucial in recovering causality ) , in a similar way as in the case presented above in this section for the usual ( magnetic ) version of the van kampen experiment . in this electric analog that we are discussing now ,
the causally - offending part of the electric aharonov - bohm phase difference will be cancelled by a magnetic type of phase , that originates from the magnetic field that is associated with the @xmath27-dependence of the electric field @xmath329 outside the cages .
we conclude that our ( exact ) results accomplish precisely what peshkin has in mind in his discussion ( on causality ) , but in a direct and fully quantitative manner , and with _ no ansatz _ based on an incorrect form .
returning for the moment to only one of the many misconceptions briefly pointed out in this paper , we should emphasize further that improper uses of simple dirac phases in the literature are not rare or marginal@xmath1 even in feynman@xcite it is stated that the simple phase factor @xmath365 is valid even for dynamic fields ; this is also explicitly stated in erlichson s
review@xcite @xmath2 silverman@xcite being the only report that we are aware of with a careful wording about ( [ wrong ] ) being only restrictedly valid ( for @xmath27-independent @xmath42 and @xmath48-independent @xmath5 ) , although even there the nonlocal terms have been missed . with respect to the presence of _ fields _ in the _ phases _ of quantum mechanical wavefunctions that we find
, it should be stressed that at the level of the basic lagrangian @xmath366 there are no fields present , and the view holds in the literature@xcite that electric fields or magnetic fields can not contribute directly to the phase .
this view originates from the path - integral treatments widely used ( where the lagrangian determines directly the phases of propagators ) , but , nevertheless , our canonical treatment shows that fields _ do _ contribute nonlocally , and they are actually crucial in recovering relativistic causality .
moreover , path - integral discussions@xcite of the van kampen case use wave ( retarded)-solutions for @xmath42 ( hence in the lorenz gauge ) and are incomplete ; our results take advantage of the retardation of _ fields _ @xmath367 and @xmath368 * * ( true in _ any _ gauge ) , and _ not _ of potentials . in addition , troudet@xcite correctly states that his path - integral treatment is good for not highly - delocalized states in space , and that in case of delocalization the proper treatment would be much more complicated , and would require a much more complete analysis .
it is fair to state that such a complete analysis has actually been provided in the present work
. it should be added that the van kampen paradox seems to be still thought of as remarkable @xcite .
the present work has provided a natural and general _ resolution _ , and most importantly , through nonlocal ( and relativistically causal ) propagation of wavefunction - phases . finally , in trying to explore an even broader significance of the new solutions
, one may wonder about possible consequences of the nonlocal terms if these are included in more general physical models that have a gauge structure ( in condensed matter or high energy physics ) .
it is also worth mentioning that if one follows the same unconventional method ( of solution of pdes ) with the maxwell s equations for the electric and magnetic fields ( rather than with the pdes of eq.([basic11 ] ) for the potentials that give @xmath25 ) , the corresponding nonlocal terms can be derived in a similar manner , and one can then see that these nonlocal terms essentially demonstrate the causal propagation of the radiation electric and magnetic fields outside physically inaccessible confined sources ( i.e. solenoids or electric cages ) .
although this is of course widely known at the level of classical fields , a major conclusion that can be drawn from the present work ( at the level of gauge transformations ) is that _ a corresponding causality also exists at the level of quantum mechanical phases _ as well , and this is enforced by the nonlocal terms in @xmath27-dependent cases .
it strongly indicates that the nonlocal terms found in this work at the level of quantum mechanical phases reflect a causal propagation of wavefunction phases * in the schrdinger picture * ( at least one part of them @xmath2 the one containing the fields @xmath2 which competes with the aharonov - bohm types of phases containing the potentials ) .
this is an entirely new concept ( given the _ local nature _ but also the
_ nonrelativistic character _ of the schrdinger equation ) and deserves to be further explored .
it would indeed be worth investigating possible applications of the above results ( of nonlocal phases of wavefunctions , solutions of the local schrdinger equation ) to @xmath27-dependent single- _ vs _ double - slit experiments recently discussed by the group of aharonov@xcite who use a completely different method , with modular variables in the heisenberg picture ( presented as the sole method appropriate for problems of this type ) .
one should also note recent work@xcite , that rightly emphasizes that physics can not currently predict how we dynamically go from the single - slit diffraction pattern to the double - slit diffraction pattern ( whether it is in a gradual and causal manner or not ) and where a relevant experiment is proposed to decide on ( address ) exactly this largely unknown issue .
application of our nonlocal terms to such questions in analogous experiments ( i.e. by introducing ( finite ) scalar potentials on slits in a @xmath27-dependent way ) provides a completely new formulation for addressing causal issues of this type , and is worth of further investigation .
furthermore , it is worth noting that , if @xmath56 s were substituted by gravitational fields and @xmath266 s by coriolis force fields arising in non - inertial ( rotating ) frames of reference , the above nonlocalities ( and their apparent causal nature ) could possibly have an interesting story to tell about quantum mechanical phase behavior in a relativistic / gravitational framework . finally , _
su(2 ) _ generalizations would be an obviously interesting extension of the above _
u(1 ) _ theory , and such generalizations are rather formally direct and not difficult to make@xmath7 an immediate physically interesting question would then be whether the new nonlocal terms might have a nontrivial impact on i.e. spin-@xmath369-states , since these terms would act asymmetrically on opposite spins ( the nonlocal @xmath196-terms being relevant for zeeman interactions , and the nonlocal @xmath56-terms possibly having a role if the above results were applied i.e. to condensed matter systems with strong spin - orbit coupling@xcite ) .
we conclude that a nonlocal and causal behavior exists at the level of quantum mechanical phases , even for solutions of the nonrelativistic ( and local ) schrdinger equation and this behavior is enforced by the nonlocal terms derived in the present work ( through the well - known causal behavior of fields ) .
our ( exact and analytical ) results accomplish precisely what peshkin has in mind in his discussion ( on causality ) of the electric aharonov - bohm effect , but in a direct and fully quantitative manner , and with _ no ansatz _ based on an incorrect form .
another pleasing characteristic of our results that , as far as we are aware , has no parallel in other formulations , is that they give a rather natural account of the dynamical nonlocality character@xcite attributed to the various aharonov - bohm phenomena ( magnetic , electric or combined ) , although @xmath2 in the present work @xmath2 this dynamical quantum nonlocality seems to simultaneously respect causality in a deterministic way , i.e. without requiring the necessity of invoking the uncertainty principle .
the nonlocal terms found in this work at the level of @xmath25 reflect a causal propagation of wavefunction - phases _ in the schrdinger picture _ , with possible immediate applications to @xmath27-dependent slit - experiments recently discussed using the heisenberg picture@xcite .
application of our nonlocal terms to such problems ( i.e. by introducing @xmath27-dependent scalar potentials on the slits ) provides a new and _ direct _ formulation for addressing causal issues of such @xmath27-dependent slit - systems .
finally , one can not refrain from wondering about the analogs of these new nonlocalities in many areas of physics where geometric or topological phases@xcite appear as the central quantities ( these always being of the form of integrals of some effective ( or emergent ) potentials ( that are determined by the system , i.e. by band structures , as in the rapidly expanding area of topological insulators@xcite ) ) .
a natural thought then arising from the present work is that , if those emerging potentials are not of a type that would correspond to zero fields ( in the physically accessible regions ) , but describe nonvanishing effective fields in ( even remote ) spacetime regions that are accessible to the particles , then one would expect that the new nonlocalities should be seriously taken into account @xmath2 these having certain dynamical consequences that may have not received an entirely proper treatment in earlier works . given the popularity and importance of these areas in the whole of physics , issues such as the above would certainly deserve further study .
graduate students kyriakos kyriakou and georgios konstantinou of the university of cyprus and dr .
areg ghazaryan of yerevan state university are acknowledged for having carefully reproduced all results .
georgios konstantinou is also gratefully acknowledged for having drawn all figures of this article .
cleopatra christoforou of the department of mathematics and statistics of the university of cyprus is acknowledged for a discussion concerning the mathematical method followed .
k. moulopoulos , nonlocal phases of local quantum mechanical wavefunctions in static and time - dependent aharonov - bohm experiments , _ _ journal of physics a : math__ematical and _ theoretical _ , vol .
35 , 2010 , article i d * * : * * 354019 , pp . 1 - 32
. j. tollaksen , y. aharonov , a. casher , t. kaufherr , and s. nussimov , quantum interference experiments , modular variables and weak measurements , _ new journal of physics _ ,
vol . 12 , 2010 , article i d * * : * * 013023 , pp . 1 - 29 .
g. p. he , flexible scheme for measuring experimentally the speed of the response of quantum states to the change of the boundary condition , http://arxiv.org/ps_cache/arxiv/pdf/0907/0907.1974v2.pdf ( nov . 2009 ) , pp . 1 - 6 .
f. g. werner & d. r. brill , significance of electromagnetic potentials in the quantum theory in the interpretation of electron interferometer fringe observations , _ _ physical review lett__ers , vol .
4 , no . 7 , 1960 , pp .
344 - 347 .
p. g. luan & c. s. tang , charged particle motion in a time - dependent flux - driven ring@xmath1 an exactly solvable model , _ journal of physics : condensed matter _ ,
19 , 2007 , article i d * * : * * 176224 , pp
. 1 - 10 .
* figure 1 . *
( color online)@xmath1 examples of simple field - configurations ( in simple - connected regions ) , where the nonlocal terms exist and are nontrivial , but can easily be determined@xmath1 ( a ) a striped case in 1 + 1 spacetime , where the electric flux enclosed in the observation rectangle is dependent on @xmath27 but independent of @xmath49 ; ( b ) a triangular distribution in 2-d space , where the part of the magnetic flux inside the corresponding observation rectangle depends on _ both
_ @xmath49 _ and _ @xmath231 . the appropriate choices for the corresponding nonlocal functions @xmath133 and @xmath151 for case ( a ) , or @xmath133 and @xmath209 for case ( b ) , are given in the text ( sections iii and iv respectively ) .
* figure 2 . *
( color online)@xmath1 the standard double - slit apparatus with an additional strip of a perpendicular magnetic field @xmath266 of width @xmath370 placed between the slit - region and the observation screen . in the text we deal for simplicity with the case
@xmath371 so that deflections ( of the semiclassical trajectories ) due to the lorentz force , shown here for a negative charge @xmath10 , are very small .
* figure 3 . *
( color online)@xmath1 the analog of fig.2 ( again for a negative @xmath10 ) but with an additional electric field parallel to the observation screen that is turned on for a time interval @xmath290 . in the text we deal for simplicity with the case @xmath372 ( with @xmath373 , @xmath285 the de broglie wavelength ) , so that deflections ( of the semiclassical trajectories ) due to the electric force are again very small . for both fig.2 and 3 ,
it is shown in the text that @xmath374 hence we observe an extra minus sign compared to what is usually reported in the literature .
* figure 4 . *
( color online)@xmath1 the analog of paths of fig.1 but now in 2 + 1 spacetime for the van kampen thought - experiment , when the instant of observation @xmath27 is so short that the physical information has not yet reached the spatial point of observation @xmath188 .
the two solutions ( that , for wavepackets , have to be subtracted in order to give the phase difference at @xmath375 ) are given in the text , and are here characterized through their electric field @xmath56-line integral behavior@xmath1 electric field path ( i ) ( the red - arrow route ) denotes solution ( [ lambdafin ] ) , and electric field path ( ii ) ( the green - arrow route ) denotes solution ( [ lambdafull2 ] ) .
note that the strips of fig.1(a ) have now given their place to a light - cone . at the point of observation ( that lies outside this light - cone )
the aharonov - bohm phase difference has now become causal due to cancellations between the two solutions ( the two electric field paths above ) . | generalized solutions of the standard gauge transformation equations are presented and discussed in physical terms .
they go beyond the usual dirac phase factors and they exhibit nonlocal quantal behavior , with the well - known relativistic causality of classical fields affecting directly the _ phases _ of wavefunctions in the schrdinger picture . these nonlocal phase behaviors , apparently overlooked in path - integral approaches , give a natural account of the dynamical nonlocality character of the various ( even static ) aharonov - bohm phenomena , while at the same time they seem to respect causality . for particles passing through nonvanishing magnetic or electric fields
they lead to cancellations of aharonov - bohm phases at the observation point , generalizing earlier semiclassical experimental observations ( of werner & brill ) to delocalized ( spread - out ) quantum states .
this leads to a correction of previously unnoticed sign - errors in the literature , and to a natural explanation of the deeper reason why certain time - dependent semiclassical arguments are consistent with static results in purely quantal aharonov - bohm configurations .
these nonlocalities also provide a remedy for misleading results propagating in the literature ( concerning an uncritical use of dirac phase factors , that persists since the time of feynman s work on path integrals ) .
they are shown to conspire in such a way as to exactly cancel the instantaneous aharonov - bohm phase and recover relativistic causality in earlier paradoxes ( such as the van kampen thought - experiment),@xmath0and to also complete peshkin s discussion of the electric aharonov - bohm effect in a _ causal _ manner .
the present formulation offers a direct way to address time - dependent single- _ vs _ double - slit experiments and the associated causal issues issues that have recently attracted attention , with respect to the inability of current theories to address them .
keywords@xmath1 aharonov - bohm , gauge transformations , dirac phase factor , quantum phases |
theories of critical phenomena in low dimensions are now quite well established due to the concepts of the renormalization group ( rg ) and universality .
the two - dimensional ( 2d ) ising model and the 2d @xmath3 model are two classical model systems that have considerably enhanced our understanding of these concepts .
while the former exhibits an order - disorder phase transition , the physics of the latter model is better described in terms of vortex - antivortex unbinding , where the kosterlitz - thouless ( kt ) picture emerges @xcite .
closely related to these models , one may also consider the 2d @xmath0-state clock model defined by the following hamiltonian : @xmath4 where @xmath5 means interaction strength , the summation is over the nearest neighbors , and @xmath6 with @xmath7 . as can be easily seen , this model has been extensively studied as a bridge between the ising model ( @xmath8 ) and the @xmath3 model ( @xmath9 ) .
but it does not mean that @xmath10 is actually required for a kt transition to be observed : rather , the kt behavior becomes possible for @xmath11 with an intermediate quasi - ordered massless phase between an ordered phase at low temperature and a disordered phase at high temperature .
a hand - waving rg argument is that domain walls get so floppy to effectively provide block spins with the continuous @xmath1 symmetry over large length scales @xcite . in order to estimate @xmath12 ,
let us generalize eq .
( [ eq : h ] ) so that @xmath13 where the spin - interaction potential @xmath14 has the @xmath15 symmetry . the villain @xmath0-state clock model chooses , for example , @xmath16 \right\},\ ] ] where @xmath17 with the boltzmann constant @xmath18 and temperature @xmath19 .
this potential form has been introduced to separate the vortex degrees of freedom from the spin - wave degrees of freedom as an approximate version of the @xmath3 model @xcite .
this villain clock model on the square lattice possesses self - duality @xcite , predicting that @xmath20 based on the following argument @xcite : suppose that there are only two phases .
then , the phase transition should occur at the dual point @xmath21 .
the villain @xmath3 model has a transition temperature around @xmath22 . for @xmath23 ,
therefore , @xmath24 becomes lower than @xmath25 .
however , this is against the correlation inequality @xcite , which says that @xmath26 where @xmath27 and @xmath28 are arbitrary spin indices and the brakets mean statistical averages of the villain @xmath0-state clock model and that of the villain @xmath3 model , respectively , at the same temperature .
it expresses an intuitively obvious idea that ordering is generally weaker in the latter model since it has a greater number of possible spin configurations , and this sort of statement holds even if we work with the cosine potential . in short , the transition temperature of the villain @xmath0-state clock model can not be lower than @xmath29 , contradicting the duality argument in the two - phase picture .
therefore , the two - phase picture is not valid and we are forced to have three phases at least for @xmath23 .
the same conclusion can be drawn for the triangular and honeycomb lattices as well @xcite .
since this argument relies on the specific choice of @xmath30 in the villain form , one might wonder if the same conclusion can be drawn for the usual clock model with the cosine potential .
for instance , a recent claim was @xmath31 based on helicity modulus @xcite , though more recent calculations have confirmed a normal kt transition for @xmath32 @xcite . since the four - state clock model exactly belongs to the ising universality class @xcite , the only remaining case to consider is @xmath33 .
although the migdal - kadanoff rg transformation predicts only two phases @xcite , there is ample evidence that this five - state clock model does have three phases @xcite , and theoretical arguments for the three - phase picture have been motivated and supported by these observations .
we first calculate the conformal charge @xmath34 from the leading finite - size correction of the free energy on semi - infinite strips of the square geometry @xcite , and this also indicates the existence of a temperature region with @xmath35 as in the six - state clock model [ compare fig .
[ fig : transfer](a ) and [ fig : transfer](b ) ] . even if not completely rigorous , a convincing argument for the three - phase picture originates from a field - theoretic construction @xcite , which shows that there exists a conformally invariant point with conformal charge @xmath36 between the potts model and the clock model , and conjectures that this is where the two critical lines merge .
this conjecture is found consistent with numerical results @xcite . in short ,
the dominant consensus is that the five - state clock model belongs to the same universality class as the villain five - state clock model , where one kt transition and its dual separate three phases . .
the vertical lines represent @xmath37 and @xmath38 obtained in ref .
( b ) the same quantity for the five - state clock model , where the vertical lines represent @xmath37 and @xmath38 obtained in ref .
the free energy of each finite - sized system is calculated by using the transfer - matrix method @xcite and the bulk free energy is obtained by the bulirsch - stoer extrapolation algorithm @xcite .
note the convergence to @xmath35 as @xmath39 grows in both the cases.,title="fig:",scaledwidth=45.0% ] .
the vertical lines represent @xmath37 and @xmath38 obtained in ref .
( b ) the same quantity for the five - state clock model , where the vertical lines represent @xmath37 and @xmath38 obtained in ref .
the free energy of each finite - sized system is calculated by using the transfer - matrix method @xcite and the bulk free energy is obtained by the bulirsch - stoer extrapolation algorithm @xcite .
note the convergence to @xmath35 as @xmath39 grows in both the cases.,title="fig:",scaledwidth=45.0% ] -state clock model .
the transition points for @xmath40 are taken from refs .
@xcite , and the solid lines are guides to eyes .
note that both the transition points for @xmath33 are above the dotted horizontal line , representing @xmath41.,scaledwidth=45.0% ] however , we have observed that helicity modulus of the five - state clock model does not vanish at the upper transition @xcite , which casts doubt on its kt nature .
that is , if the system underwent a kt transition , this quantity would show a jump to zero @xcite . here , the helicity modulus is defined as response to an infinitesimal twist @xmath42 across the system in the @xmath43 direction : @xmath44 where @xmath45 and @xmath46 are summed over all the links in the @xmath43 direction .
although extensive monte carlo calculations have shown that all the other features of the transition are not inconsistent with the kt picture @xcite , these calculations raise further questions since the transition temperatures are estimated as @xmath47 and @xmath48 , respectively : estimates of the transition temperature @xmath41 of the @xmath3 model vary among authors @xcite , but one can safely say @xmath49 .
the point is that @xmath41 is still well below the lower transition temperature of the five - state clock model , as well as the dual point @xmath50
@xcite ( see fig . [
fig : phase ] ) .
this implies a sort of _ prematurity _ in the emergence of three phases in the five - state clock model in that the duality can not enforce it as in the villain clock model .
in addition , when the mass gap in the hamiltonian formulation has a singularity of the following form : @xmath51 where @xmath52 is the reduced temperature and @xmath53 is a constant , the kt prediction of @xmath54 @xcite appears consistent only when @xmath2 , whereas it converges to @xmath55 at @xmath33 @xcite . note that a second - order phase transition would be represented by @xmath56
all these signatures suggest something inherently different in the five - state clock model from the six - state clock model as well as from the four - state clock model . to sum up , the situation is that the three - phase picture looks convincing in the five - state clock model , but that consistency with the kt picture is still doubtful . in this work ,
we explicitly show that the nonvanishing helicity modulus of the five - state clock model means a residual five - fold symmetry , which is not completely washed away by spin fluctuations even at @xmath57 .
it implies a possibility that the continuous @xmath1 symmetry as in the @xmath3 model sets in only when @xmath58 , which means a difference from the villain formulation . after presenting our numerical results in sec .
[ sec : ftbc ] , we discuss a theoretical attempt to explain this feature by introducing the second length scale of vortex composites in sec .
[ sec : discuss ] , and then conclude this work .
in the fluctuating twist boundary condition ( ftbc ) @xcite , there exists interaction among the boundary spins as in the periodic boundary condition .
a difference is that here we apply a gauge field that adds twist @xmath42 in the interaction potential .
if @xmath42 is fixed at zero , it is identical to the periodic boundary condition , whereas if it is fixed at @xmath59 , it corresponds to the anti - periodic boundary condition .
as the name indicates , the ftbc lets the twist @xmath42 also fluctuate in time following the metropolis update rule .
the free energy is written as @xmath60 , and the probability to observe @xmath42 at a given @xmath61 will be @xmath62 $ ] @xcite . in this way , we measure how the free energy changes as @xmath42 varies .
when @xmath42 is a continuous variable , since @xmath63 , this method provides an estimate of the helicity modulus defined in eq .
( [ eq : y ] ) .
let us begin with the 2d ising model ( @xmath8 ) on the square lattice .
since @xmath64 is identically zero , its helicity modulus @xmath65 is proportional to the internal energy density , which obviously does not vanish at any temperature [ see eq .
( [ eq : y ] ) ] . on the other hand ,
the free - energy difference between the periodic boundary condition and the anti - periodic boundary condition becomes exponentially small as the system size grows if the temperature is higher than @xmath66 in units of @xmath67 .
these two facts immediately suggest that the free energy @xmath68 has a smooth barrier between @xmath69 and @xmath59 and another between @xmath70 and @xmath71 by symmetry .
this speculation is directly confirmed in fig .
[ fig : ftbc](a ) obtained by using the ftbc , where we see that the system is disordered but preserves its up - down symmetry .
it is also intuitively clear that the system begins to prefer @xmath70 to @xmath72 as @xmath19 passes @xmath73 from above so that @xmath68 changes its shape as shown in fig .
[ fig : ftbc](a ) . to understand the nonvanishing @xmath74 on physical grounds , it is instructive to recall that the free energy is written as @xmath75 with internal energy @xmath76 and entropy @xmath77 . since the first term in eq .
( [ eq : y ] ) is related to the second derivative of @xmath78 , the second term describing fluctuations should be related to the entropic contribution to the free - energy change .
we can say that if @xmath42 is very small , it is difficult for the system to excite domain walls whose entropy could compensate for the energy change . in this sense , the nonvanishing @xmath74 means that enough domain walls , a discrete version of spin waves , are not generated to cancel out the external perturbation @xmath42 , possibly due to the finite energy scale needed to excite them . where @xmath79 is probability of observing twist @xmath42 under the fluctuating twist boundary condition at temperature @xmath19 .
in practice , the twist @xmath42 is given as a @xmath80-state clock spin to simulate a continuous variable .
( a ) the discrete symmetry is clearly detected in the disordered phase of the ising model , and ( b ) the same statement can be made for the four - state clock model .
( c ) on the other hand , the discrete symmetry changes to the continuous @xmath1 symmetry in the six - state clock model and ( d ) manifests itself only at much lower @xmath19 .
the system size is taken as @xmath81 in every case.,title="fig:",scaledwidth=45.0% ] where @xmath79 is probability of observing twist @xmath42 under the fluctuating twist boundary condition at temperature @xmath19 . in practice , the twist @xmath42 is given as a @xmath80-state clock spin to simulate a continuous variable .
( a ) the discrete symmetry is clearly detected in the disordered phase of the ising model , and ( b ) the same statement can be made for the four - state clock model .
( c ) on the other hand , the discrete symmetry changes to the continuous @xmath1 symmetry in the six - state clock model and ( d ) manifests itself only at much lower @xmath19 .
the system size is taken as @xmath81 in every case.,title="fig:",scaledwidth=45.0% ] where @xmath79 is probability of observing twist @xmath42 under the fluctuating twist boundary condition at temperature @xmath19 . in practice , the twist @xmath42 is given as a @xmath80-state clock spin to simulate a continuous variable .
( a ) the discrete symmetry is clearly detected in the disordered phase of the ising model , and ( b ) the same statement can be made for the four - state clock model .
( c ) on the other hand , the discrete symmetry changes to the continuous @xmath1 symmetry in the six - state clock model and ( d ) manifests itself only at much lower @xmath19 .
the system size is taken as @xmath81 in every case.,title="fig:",scaledwidth=45.0% ] where @xmath79 is probability of observing twist @xmath42 under the fluctuating twist boundary condition at temperature @xmath19 . in practice , the twist @xmath42 is given as a @xmath80-state clock spin to simulate a continuous variable .
( a ) the discrete symmetry is clearly detected in the disordered phase of the ising model , and ( b ) the same statement can be made for the four - state clock model .
( c ) on the other hand , the discrete symmetry changes to the continuous @xmath1 symmetry in the six - state clock model and ( d ) manifests itself only at much lower @xmath19 .
the system size is taken as @xmath81 in every case.,title="fig:",scaledwidth=45.0% ] the four - state clock model with interaction strength @xmath5 has the same partition function as two independent ising systems with interaction strength @xmath82 @xcite .
this model has @xmath35 since each ising model carries conformal charge @xmath83
@xcite , but the massless phase exists only at a single point so the model has only two phases separated at @xmath84 .
it is also obvious from this equivalence that @xmath74 does not vanish at @xmath85 , and we see how the system preserves its four - fold symmetry in fig .
[ fig : ftbc](b ) . in terms of vortices ,
this is related to the fact that a vortex is not a really independent object but merely results from two independent ising domain walls that happen to cross at a point . in the six - state clock model , on the other hand , we expect from the vanishing @xmath74 that the six - fold symmetry will be manifested only at @xmath86 and is not visible at higher temperatures . in the language of the rg
, it means that the system can respond to perturbations as if it were a continuous spin system on average over large length scales .
our numerical calculations on the @xmath87 square lattice clearly support this picture as depicted in figs .
[ fig : ftbc](c ) and [ fig : ftbc](d ) .
note that these results are obtained with quite a small size , i.e. , @xmath81 , so that the transition temperature @xmath38 tends to be overestimated [ see , e.g. , figs .
[ fig : five](c ) or [ fig : five](d ) ] .
nevertheless , these give us a qualitatively correct picture that remains true for larger system sizes ( see ref .
@xcite , which shows data up to @xmath88 ) . in short
, the six - state clock model effectively exhibits the continuous @xmath1 symmetry whereby the genuine kt behavior becomes possible .
of the villain five - state clock model [ cf .
[ fig : ftbc](c ) ] .
( b ) five - fold symmetry is clearly visible in the five - state clock model with the cosine potential .
we take @xmath81 as the system size in both the cases .
( c ) as the system size grows in the villain five - state clock model , the change in curvature at @xmath89 [ i.e. , eq .
( [ eq : y ] ) ] becomes sharper , approaching the universal jump to zero .
the vertical line shows the critical temperature @xmath90 estimated by extrapolating the size dependence of the universal jump condition @xmath91 @xcite according to the kt scenario @xmath92 .
( d ) when @xmath93 increases in the five - state clock model , the curvature at @xmath89 remains finite even above @xmath94 @xcite , which is represented as the vertical line.,title="fig:",scaledwidth=45.0% ] of the villain five - state clock model [ cf .
[ fig : ftbc](c ) ] .
( b ) five - fold symmetry is clearly visible in the five - state clock model with the cosine potential .
we take @xmath81 as the system size in both the cases .
( c ) as the system size grows in the villain five - state clock model , the change in curvature at @xmath89 [ i.e. , eq .
( [ eq : y ] ) ] becomes sharper , approaching the universal jump to zero .
the vertical line shows the critical temperature @xmath90 estimated by extrapolating the size dependence of the universal jump condition @xmath91 @xcite according to the kt scenario @xmath92 .
( d ) when @xmath93 increases in the five - state clock model , the curvature at @xmath89 remains finite even above @xmath94 @xcite , which is represented as the vertical line.,title="fig:",scaledwidth=45.0% ] of the villain five - state clock model [ cf .
[ fig : ftbc](c ) ] .
( b ) five - fold symmetry is clearly visible in the five - state clock model with the cosine potential .
we take @xmath81 as the system size in both the cases .
( c ) as the system size grows in the villain five - state clock model , the change in curvature at @xmath89 [ i.e. , eq .
( [ eq : y ] ) ] becomes sharper , approaching the universal jump to zero .
the vertical line shows the critical temperature @xmath90 estimated by extrapolating the size dependence of the universal jump condition @xmath91 @xcite according to the kt scenario @xmath92 .
( d ) when @xmath93 increases in the five - state clock model , the curvature at @xmath89 remains finite even above @xmath94 @xcite , which is represented as the vertical line.,title="fig:",scaledwidth=45.0% ] of the villain five - state clock model [ cf .
[ fig : ftbc](c ) ] .
( b ) five - fold symmetry is clearly visible in the five - state clock model with the cosine potential .
we take @xmath81 as the system size in both the cases .
( c ) as the system size grows in the villain five - state clock model , the change in curvature at @xmath89 [ i.e. , eq .
( [ eq : y ] ) ] becomes sharper , approaching the universal jump to zero .
the vertical line shows the critical temperature @xmath90 estimated by extrapolating the size dependence of the universal jump condition @xmath91 @xcite according to the kt scenario @xmath92 .
( d ) when @xmath93 increases in the five - state clock model , the curvature at @xmath89 remains finite even above @xmath94 @xcite , which is represented as the vertical line.,title="fig:",scaledwidth=45.0% ] it is interesting to see that the villain five - state clock model also shows the same kind of behavior , supporting theoretical works based on the villain formulation [ fig .
[ fig : five](a ) ] .
however , when the method is applied to the conventional five - state clock model to obtain @xmath68 , we see that the behavior is more similar to what we observed for @xmath8 or @xmath95 than to @xmath96 , in that the model exhibits its five - fold symmetry even in the disordered phase [ fig .
[ fig : five](b ) ] .
note that the free energy @xmath68 has a curvature near @xmath70 , which corresponds to the nonvanishing @xmath74 .
therefore , the nonvanishing @xmath74 signals the residual discrete symmetry above the upper transition , which remains there even if the system size grows [ compare fig .
[ fig : five](c ) and [ fig : five](d ) ] , so we can conclude that the upper phase transition of the five - state clock model is not of the conventional kt type . and ( b ) its scaling collapse where the exact values @xmath97 and @xmath98 are used .
the horizontal line represents @xmath99 .
( c ) interfacial free energy of the five - state clock model , where the two vertical lines represent @xmath37 and @xmath38 measured in ref .
@xcite , respectively .
( d ) scaling collapse attempted with the dual point @xmath50
@xcite and a trial exponent @xmath100 used in ref .
error bars are usually smaller than the symbol sizes.,title="fig:",scaledwidth=45.0% ] and ( b ) its scaling collapse where the exact values @xmath97 and @xmath98 are used .
the horizontal line represents @xmath99 .
( c ) interfacial free energy of the five - state clock model , where the two vertical lines represent @xmath37 and @xmath38 measured in ref .
@xcite , respectively .
( d ) scaling collapse attempted with the dual point @xmath50
@xcite and a trial exponent @xmath100 used in ref .
error bars are usually smaller than the symbol sizes.,title="fig:",scaledwidth=45.0% ] and ( b ) its scaling collapse where the exact values @xmath97 and @xmath98 are used .
the horizontal line represents @xmath99 .
( c ) interfacial free energy of the five - state clock model , where the two vertical lines represent @xmath37 and @xmath38 measured in ref .
@xcite , respectively .
( d ) scaling collapse attempted with the dual point @xmath50 @xcite and a trial exponent @xmath100 used in ref .
error bars are usually smaller than the symbol sizes.,title="fig:",scaledwidth=45.0% ] and ( b ) its scaling collapse where the exact values @xmath97 and @xmath98 are used .
the horizontal line represents @xmath99 .
( c ) interfacial free energy of the five - state clock model , where the two vertical lines represent @xmath37 and @xmath38 measured in ref .
@xcite , respectively .
( d ) scaling collapse attempted with the dual point @xmath50
@xcite and a trial exponent @xmath100 used in ref .
error bars are usually smaller than the symbol sizes.,title="fig:",scaledwidth=45.0% ] as a brief sideline , we may consider what one would find if @xmath42 also had the discrete @xmath0-fold symmetry .
for the ising model , this free - energy difference between the periodic boundary condition ( @xmath89 ) and the anti - periodic boundary condition ( @xmath101 ) is called the interfacial free energy .
this quantity is known to have the following scaling form @xcite : @xmath102 where @xmath103 is a scaling function with a universal amplitude @xmath104 and @xmath105 is the scaling exponent for the divergence of the correlation length @xmath106 .
it is straightforward to generalize the notion of the interfacial free energy to the case of the @xmath0-state clock model by measuring the free - energy difference caused by setting @xmath107 . in other words ,
we define @xmath108 , and this quantity shows how the lower envelope of @xmath68 changes its shape .
note that the envelope is not distinguished from @xmath68 if @xmath2 so that we do not have to consider such cases separately .
for the four - state clock model , we observe that this method indeed yields correct scaling results [ figs .
[ fig : interface](a ) and [ fig : interface](b ) ] . when applied to the five - state clock model
, @xmath109 does vanish in the high - temperature phase by construction in contrast to the helicity modulus [ fig .
[ fig : interface](c ) ] .
the question is then how this quantity describes the critical properties . in ref .
@xcite , the authors attempt scaling collapse with @xmath110 under the assumption that the model undergoes a usual continuous phase transition . using this trial exponent
, we also observe reasonable scaling collapse around the dual temperature @xmath50 @xcite as depicted in fig .
[ fig : interface](d ) , but the possibility of a crossover phenomenon can not be ruled out for such small sizes .
a possible reason for the peculiarity of the five - state clock model has already been argued in refs .
this theory points out a possibility that the model may have two different length scales , one from individual vortices and the other from composites of vortices . according to this scenario ,
the nonvanishing helicity modulus is a finite - size effect governed by the larger length scale of the composites , which will eventually become negligible in the thermodynamic limit .
the correlation due to the composites is destroyed at a higher temperature called the disorder line , roughly estimated as @xmath111 \approx 1.81 $ ] , where the correlated liquid becomes a potts gas .
it turns out hard to examine the existence of the second length scale with numerical calculations : if there really was a second length scale , it would have to get shorter with increasing @xmath19 so that the helicity modulus would then be more sensitive to the lattice size . for this reason ,
we have tried to observe the helicity modulus as a function of @xmath93 with fixing the temperature at @xmath112 well above @xmath113 .
however , it appears to converge to a nonzero value at least up to @xmath114 ( see also ref .
@xcite ) , implying that the second length scale , if it exists , might be still greater than this size at such high @xmath19 .
we additionally note that ref .
@xcite suggests a _ conspiracy _ between the clock number @xmath33 and the four - fold symmetry of the square lattice : for a unit - charge vortex located at a square plaquette , the sum of angle differences around it is @xmath115 , and there are only _ four _ angle differences , each of which can take an integer multiple of @xmath116 .
it means that there should be at least one angle difference greater than @xmath116 , likely to be @xmath117 .
so the argument for the origin of the composite on the square lattice is that such a high - energy domain wall tends to appear as a double strand of two single domain walls in the vicinity of the vortex - antivortex unbinding transition , forming a correlated local structure ( see ref .
@xcite for related discussions ) . however , it is not easy to extend such an argument to other 2d lattices as already noted in ref .
@xcite , and it turns out that we need a more universal explanation because the nonvanishing helicity modulus is observed on the triangular and the honeycomb lattices as well ( fig .
[ fig : q5 ] ) .
therefore , our observation is only partially consistent with the theory in refs .
@xcite at the moment , and it seems quite nontrivial to explain how the discrete five - fold symmetry emerges in vortex composites over more than @xmath118 lattice spacings . if we once again borrow the hand - waving rg argument in sec .
[ sec : intro ] @xcite , our observation rather implies another possibility that the discreteness in the spin degree of freedom is not completely renormalized away with the cosine potential and that this may survive further rg transformations . ) ] of the five - state clock model ( a ) on the triangular lattice and ( b ) on the honeycomb lattice .
error bars are not larger than the symbol sizes .
see also ref .
@xcite about computing @xmath74 on these lattices.,title="fig:",scaledwidth=45.0% ] ) ] of the five - state clock model ( a ) on the triangular lattice and ( b ) on the honeycomb lattice .
error bars are not larger than the symbol sizes .
see also ref .
@xcite about computing @xmath74 on these lattices.,title="fig:",scaledwidth=45.0% ] in summary , we have found the origin of the nonvanishing helicity modulus numerically observed in the five - state clock model : it originates from the five - fold symmetry , which is not replaced by the continuous @xmath1 symmetry even deep inside the high - temperature phase and clearly manifests itself under the ftbc . as a consequence , although our transfer - matrix calculation supports the three - phase picture for the five - state clock model , the nature of the transition is not fully developed to the genuine kt type , differently from its villain - approximated version , in our monte carlo calculations .
since there is not enough numerical evidence for a vanishing helicity modulus in the five - state clock model , it is only for @xmath2 that one can be sure of the three phases separated by a genuine kt transition and its dual , and we suggest that the transition of the five - state clock model is a weaker cousin of the kt transition .
is thankful to h. park for bringing his attention to ref .
conversations with m. den nijs and j. m. kosterlitz are also gratefully acknowledged .
was supported by the academy of finland through its centres of excellence program ( project no .
b.j.k . was supported by the national research foundation of korea ( nrf ) grant funded by the korean government ( mest ) ( grant no . 2011 - 0015731 ) .
this work was supported by the supercomputing center / korea institute of science and technology information with supercomputing resources including technical support ( project no .
ksc-2012-c1 - 05 ) .
43ifxundefined [ 1 ] ifx#1 ifnum [ 1 ]
# 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | it is well - known that the @xmath0-state clock model can exhibit a kosterlitz - thouless ( kt ) transition if @xmath0 is equal to or greater than a certain threshold , which has been believed to be five .
however , recent numerical studies indicate that helicity modulus does not vanish in the high - temperature phase of the five - state clock model as predicted by the kt scenario . by performing monte carlo calculations under the fluctuating twist boundary condition
, we show that it is because the five - state clock model does not have the fully continuous @xmath1 symmetry even in the high - temperature phase while the six - state clock model does .
we suggest that the upper transition of the five - state clock model is actually a weaker cousin of the kt transition so that it is @xmath2 that exhibits the genuine kt behavior . |
stigler @xcite developed an asymptotic result for the trimmed mean without requiring continuity of the underlying distribution function associated with the observations .
this result was extended to non - degenerate @xmath0-statistics based on trimmed samples in borovskikh and weber @xcite .
an alternative method for developing robust versions of @xmath0-statistics is to consider the statistic formed by trimming the kernel values , rather than the observations upon which the statistic is based .
this idea is discussed in , for example , serfling @xcite , choudhury and serfling @xcite and gijbels , janssen and veraverbeke @xcite . in this paper
, we use the generalized @xmath2-statistic representation developed in serfling @xcite to obtain an asymptotic result for trimmed @xmath0-statistics under quite general conditions .
we will not require continuity of the relevant , associated distribution at the truncation points .
let @xmath3 be independent identically distributed random variables , taking values in a measurable space @xmath4 and having common distribution @xmath5 let @xmath6 be a symmetric function from @xmath7 to @xmath8 and denote by @xmath9 the right - continuous distribution function of the random variable @xmath10 .
set @xmath11 and let @xmath12 be an enumeration of the values of @xmath13 taken over the @xmath14 @xmath15-tuples in @xmath16 note that these random variables @xmath17 are , in general , dependent .
let @xmath18 denote the ordered values of @xmath19 .
the original @xmath0-statistic is defined as an average taken over the @xmath20 possible outcomes @xmath21 , that is , @xmath22 where the empirical distribution function @xmath23 of @xmath0-statistical structure is defined by @xmath24 and @xmath25 denotes the indicator of the set @xmath26 . for any @xmath27 ,
let @xmath28 $ ] , where @xmath29 $ ] denotes the largest integer less than or equal to @xmath30 if @xmath31 , then put @xmath32 the trimmed versions of @xmath0 are based on trimming the second sum in ( 1 ) , @xmath33 or on trimming of the range of integration in ( [ for1 ] ) , @xmath34 with @xmath35 and @xmath36 where @xmath37 , \gamma= \alpha,\beta.$ ] for the results that follow , it is important to note that the lower bound for the integral in ( [ for4 ] ) is included and the upper bound excluded .
this is critical since @xmath38 is a step function . with this constraint ,
we are able to obtain the asymptotic distribution of @xmath39 without imposing any conditions on the nature of @xmath40 in lemma [ lem2.3 ] , we show that @xmath41 thus , @xmath42 and @xmath43 differ in terms of their divisors , and there are possible subtle differences in the number of summands . a class of generalized @xmath2-statistics , which includes ( [ for3 ] ) and ( [ for4 ] ) , was introduced by serfling @xcite .
the trimmed @xmath0-statistics ( [ for3 ] ) and ( [ for4 ] ) are directly connected with generalized lorenz curves , which are important in financial mathematics ( see , for example , goldie @xcite , helmers and zitikis @xcite ) . clearly , @xmath44 is an unbiased estimator of @xmath45 in the case @xmath46 and @xmath47 @xmath38 reduces to the usual empirical distribution function . define the left - continuous quantile function @xmath48 for any distribution function @xmath40 the empirical quantile function @xmath49 has the form @xmath50 a large number of authors have studied the weak convergence of such @xmath2-statistics in the case @xmath51 a partial list consists of chernoff _ et al . _
@xcite , bickel @xcite , shorack @xcite , stigler @xcite , csrgo _ et al . _
@xcite , griffin and pruitt @xcite , cheng @xcite , mason and shorack @xcite . for @xmath52 , under various sets of regularity conditions , asymptotic normality of various types of generalized @xmath2-statistics has been investigated by silverman @xcite , serfling @xcite , akritas @xcite , janssen _
et al . _
@xcite , helmers and ruymgaart @xcite , gijbels _ et al .
_ @xcite and hssjer @xcite . in the aforementioned papers , for @xmath53
, the results always assumed that @xmath54 is continuous or smooth .
however , in modern statistical robust procedures and for bootstrap procedures , results allowing for the discontinuity of the underlying distribution function @xmath54 are needed .
we study the asymptotic behavior of @xmath55 and @xmath56 for any @xmath57 without imposing the requirement of continuity .
the conditions of our theorem and the limit random variable are defined via the values of quantile function @xmath58 at the points @xmath59 and @xmath60 existing results handle the cases where @xmath61 our main result is derived without this assumption of continuity .
we represent the trimmed @xmath0-statistic as a sum of classical @xmath0-statistics with bounded , non - degenerate kernels plus some smaller terms and then we apply standard results to such statistics . for convenience , in what follows , for the distribution function @xmath62
we denote the smallest quantile @xmath63 and the largest quantile @xmath64 as , respectively , @xmath65 and @xmath66 with @xmath67 .
let @xmath68 note that @xmath69 and @xmath70 are valid for all @xmath71 and the following events coincide : @xmath72 introduce the functional @xmath73 , where @xmath74\,\mathrm{d}h_f(x)\ ] ] and the following functions with @xmath75 : @xmath76 \nonumber\\ & & { } - \bigl [ ei\ { h(x , x_2,\ldots , x_m ) < \xi^+_\alpha\}\bigl(h(x , x_2,\ldots , x_m ) - \xi^+_\alpha\bigr)+\alpha\xi^+_\alpha\bigr ] -\theta,\nonumber\\ g_\alpha(x ) & = & ei\{h(x , x_2,\ldots , x_m ) < \xi^+_\alpha\}-\theta _ \alpha , \qquad \theta_\alpha = h_f(\xi^+_\alpha-),\\ g_\beta(x ) & = & ei\{h(x , x_2,\ldots , x_m ) \leq\xi^-_\beta\}-\theta _ \beta \nonumber\\ & = & 1-\theta_\beta- ei\{h(x , x_2,\ldots , x_m ) > \xi^-_\beta\},\qquad \theta_\beta = h_f(\xi^-_\beta).\nonumber\end{aligned}\ ] ] note that for all @xmath77 and @xmath75 , we have @xmath78 let @xmath79 and @xmath80 [ teo1.1 ] if @xmath81 then for any underlying distribution function @xmath82 , we have @xmath83 where @xmath84 is a trivariate gaussian random vector with mean vector zero and covariance matrix @xmath85 [ cor1.2 ] for any underlying distribution function @xmath54 , we have , when @xmath86 , @xmath87 where @xmath84 is a trivariate gaussian random vector defined as in theorem [ teo1.1 ] . [ cor1.3 ] suppose that the quantile function @xmath88 is continuous at the points @xmath59 and @xmath60 if @xmath89 then @xmath90 for the simple case @xmath91 , the functions in ( [ for7 ] ) reduce to @xmath92 a useful application of the theorem for the @xmath93 case is for the kernel @xmath94 @xmath95 this provides the asymptotic behavior of a natural , alternative robust version of the sample variance .
we will now develop explicit expressions for the terms in a more interesting example .
let @xmath96 with @xmath52 .
let @xmath97 be the distribution function of @xmath98 and let @xmath99 then @xmath100 and @xmath101}y\,\mathrm{d}h_f(y ) \nonumber\\ & = & i\{\xi^+_\alpha\leq x \leq\xi^-_\beta\ } x ( f(x))^{m-1 } - \int_{[\xi^+_\alpha,\xi^-_\beta ] } y ( f(y))^{m-1 } \,\mathrm{d } f(y ) \nonumber \\ & & { } + \int_{[\xi^+_\alpha,\xi^-_\beta ] } \bigl(i\{x < y\ } - f(y-)\bigr)y \,\mathrm{d}(f(y))^{m-1 } , \nonumber\\ g_\alpha(x ) & = & i\{x < \xi^+_\alpha\}(f(\xi^+_\alpha-))^{m-1 } - ( f(\xi^+_\alpha-))^{m } , \nonumber\\ g_\beta(x ) & = & i\{x \leq \xi^-_\beta\}(f(\xi^-_\beta))^{m-1}-(f(\xi^-_\beta))^{m}.\end{aligned}\ ] ] in addition , @xmath102 ^ 2\\ & & { } + 2eg_{\alpha\beta}(x ) [ \xi^+_\alpha g_\alpha(x ) - \xi^-_\beta g_\beta(x)],\\ \sigma_{g_\alpha}^2 & = & ( f(\xi^+_\alpha-))^{2m-1}\bigl(1 - f(\xi^+_\alpha-)\bigr),\qquad \sigma_{g_\beta}^2 = ( f(\xi^-_\beta))^{2m-1}\bigl(1 - f(\xi ^-_\beta)\bigr ) , \\ eg_{\alpha}(x)g_{\beta}(x ) & = & ( f(\xi^+_\alpha-)f(\xi^-_\beta ) ) ^{m-1 } f(\xi^+_\alpha- ) \bigl(1 - f(\xi^-_\beta)\bigr ) , \\
eg_{\alpha\beta}(x)g_{\alpha}(x ) & = & ( f(\xi^+_\alpha-))^{m } \int _ { [ \xi^+_\alpha,\xi^-_\beta ] } \bigl(1 - f(y-)\bigr)y\ , \mathrm{d}(f(y))^{m-1 } \nonumber \\ & & { } - ( f(\xi^+_\alpha-))^{m } \int_{[\xi^+_\alpha,\xi^-_\beta ] } y ( f(y))^{m-1}\ , \mathrm{d } f(y),\\ eg_{\alpha\beta}(x)g_{\beta}(x ) & = & ( f(\xi^-_\beta ) ) ^{m-1}\bigl(1-f(\xi^-_\beta)\bigr ) \int_{[\xi^+_\alpha,\xi^-_\beta ] } y ( f(y))^{m-1}\ , \mathrm{d } f(y ) \nonumber \\ & & { } + ( f(\xi^-_\beta))^{m-1}\bigl(1-f(\xi^-_\beta)\bigr ) \int_{[\xi^+_\alpha,\xi^-_\beta ] } f(y- ) y\ , \mathrm{d}(f(y))^{m-1}.\end{aligned}\ ] ] consider the distribution function @xmath103 then @xmath104 , \qquad \sigma^2_g > 0\end{aligned}\ ] ] and the limiting behavior is given by @xmath105 however , for the simpler distribution function @xmath106 we have @xmath107 , \qquad \sigma^2_g > 0\end{aligned}\ ] ] and we get the asymptotic behavior covered by janssen _ et al . _ @xcite , @xmath108
the following two lemmas are key results for the proof . for @xmath114 , write @xmath115 since @xmath116 ,
@xmath117 = @xmath118 and , by ( [ for6 ] ) , @xmath119 we can write @xmath120\\[-8pt ] & & { } - \delta\xi_\beta i\{n^-_\beta < n_\beta\}(n^-_\beta - n_\beta ) .\end{aligned}\ ] ] note that @xmath121 and @xmath122 from ( [ for6 ] ) , we have @xmath123 @xmath124 hence , in ( [ for9 ] ) , @xmath125 equation ( [ for8 ] ) follows from ( [ for9 ] ) and ( [ for10 ] ) .
this proves lemma [ lem2.1 ] .
we shall estimate @xmath129 and @xmath130 taking into account the values of the distribution function @xmath131 at @xmath132 with @xmath133 figures [ fig1 ] and [ fig2 ] illustrate the different situations that need to be considered .
* estimating @xmath142*. noting that @xmath143 , we write @xmath144 it is clear that if @xmath145 , then @xmath146 a.s .
let @xmath147 , as is the case in fig .
[ fig2 ] . in this case ,
@xmath148 and we can write @xmath149 since @xmath150 hence , we always have the relation a.s .
to estimate @xmath151 , we note first that if @xmath152 , then the indicator @xmath153 for all @xmath154 therefore , we have the inequalities @xmath155\\[-8pt ] & \leq & i\{n_\alpha < n^-_\alpha\}(n^-_\alpha - n_\alpha ) ( \xi^-_\alpha- h_{nn_\alpha})i\{\xi^-_\alpha\geq h_{nn_\alpha } \}.%\nonumber\end{aligned}\ ] ]
further , we shall apply the technique used in smirnov @xcite with a probability inequality from hoeffding @xcite ( or see , for example , serfling @xcite , pages 75 and 201 ) .
thus , @xmath156 with some positive constants @xmath157 and @xmath158 depending only on @xmath159 and @xmath160 further , @xmath161 @xmath162 for any small values of @xmath163 , by the definition of the smallest @xmath164-quantile @xmath165 under the conditions of the lemma , @xmath166 as @xmath167 hence , @xmath168 in probability as @xmath169 next , we consider @xmath170 by definition , @xmath171 and since @xmath172 and @xmath173 , it follows that the indicator @xmath174 for @xmath175 and we can write @xmath176 in ( [ for13 ] ) , we need to consider two cases : @xmath177 and @xmath178 in the first case , @xmath179 and we have the weak convergence @xmath180 as @xmath181 and the following estimates which are similar to ( [ for12 ] ) : @xmath182 where @xmath183 in addition , @xmath184 for any small values of @xmath163 because of the definition of the largest @xmath59-quantile @xmath185 hence , in the case @xmath179 , we have @xmath186 in probability as @xmath169 in the second case in ( [ for13 ] ) , @xmath187 and we have the representation @xmath188 where @xmath189 and @xmath190 ) = \mathrm{o}(n^{-1/2 } ) $ ] as , but the positive term @xmath191 is unbounded . therefore , in this case , we shall apply the estimate ( [ for14 ] ) with @xmath192 instead of @xmath193 that is , @xmath194 @xmath195 @xmath196 since the distribution function @xmath54 is continuous from the right at the point @xmath197 it follows that @xmath198 for any small @xmath199 and sufficiently large @xmath200 hence , in the second case , @xmath201 and ( [ for14 ] ) is replaced by the inequality @xmath202 which provides the desired convergence @xmath203 in probability as @xmath167 thus , we have proven that @xmath204 in probability as @xmath167 * estimating @xmath205*. noting that @xmath206 , we write @xmath207 here , by analogy with ( [ for12 ] ) , we have @xmath208\\[-8pt ] & & \quad = p\ { h_n(\xi^-_\beta-\varepsilon ) - h(\xi^-_\beta-\varepsilon ) \geq
n^{-1}n_\beta- h(\xi^-_\beta-\varepsilon)\ } \nonumber\\ & & \quad \leq c_1 \exp\ { -c_2
n \theta^2_\beta(\xi^-_\beta,\varepsilon ) \}%\nonumber\end{aligned}\ ] ] with @xmath209 and by analogy with ( [ for15 ] ) , @xmath210 here , we need to consider two cases : @xmath211 and @xmath212 in the first case , we apply the inequality ( [ for17 ] ) with sufficiently small @xmath213 in the second case , we use ( [ for17 ] ) again , but with parameter @xmath214 , as in ( [ for16 ] ) , to get @xmath215 since the distribution function @xmath54 has a limit from the left at the point @xmath216 and @xmath217 . in this result , we have @xmath218 in probability as @xmath167 finally , we consider @xmath219 since @xmath220 , we write @xmath221 & = & - i\{n^-_\beta < n_\beta\}\sum_{i = n^-_\beta+1}^{n_\beta } ( \xi^+_\beta- h_{ni})i\{\xi^-_\beta < h_{ni } < \xi^+_\beta\ } i\ { n^-_\beta < i \leq\dot n^+_\beta\}\nonumber\\[-1pt ] & & { } + i\{n^-_\beta
< n_\beta\}\sum_{i = n^+_\beta+1}^{n_\beta}(h_{ni } - \xi^+_\beta)i\{h_{ni } > \xi^+_\beta\ } i\ { i > \dot n^+_\beta\ } \nonumber\\[-1pt ] & = & -\bar j^-_\beta+ \bar j^+_\beta.\end{aligned}\ ] ] if @xmath222 , then @xmath223 a.s .
now , assume that @xmath224 in this case , @xmath225 and we have @xmath226 since @xmath227 hence , we always have @xmath223 a.s . to estimate @xmath228 ,
we write @xmath229 and apply the estimates ( [ for13])([for16 ] ) with @xmath230 instead of @xmath231 we have @xmath232 in probability as @xmath233 and hence @xmath234 in probability as @xmath167 proof of theorem [ teo1.1 ] let @xmath235 be a @xmath236-statistic of the form ( [ for1 ] ) with the kernel @xmath237 \nonumber\\ & & { } - \bigl [ i\ { h(x_1,\ldots , x_m ) < \xi^+_\alpha\}\bigl(h(x_1,\ldots , x_m ) - \xi^+_\alpha\bigr)+\alpha\xi^+_\alpha\bigr ] .
\nonumber\end{aligned}\ ] ] we see that @xmath238 it is not difficult to verify for this function that @xmath239 and @xmath240 ; in addition , @xmath241 , by the condition of the theorem .
hence , the kernel @xmath242 is non - degenerate and , by the central limit theorem for @xmath236-statistics with such bounded kernels , we have the weak convergence @xmath243 as @xmath181 ( see , for example , borovskikh @xcite ) . by the same central limit theorem , we have @xmath244 and @xmath245 as @xmath167 under the conditions of the theorem , we have @xmath246 if @xmath247 ( in this case , @xmath248 ) and @xmath249 if @xmath250 ( in this case , @xmath251 ) .
further , it is easy to prove that the covariances @xmath252 as @xmath181 , where @xmath253 now , apply lemma [ lem2.2 ] to complete the proof of theorem [ teo1.1 ] .
choudhury , j. and serfling , r.j .
generalized order statistics , bahadur representations and sequential nonparametric fixed - width confidence intervals .
_ j. statist .
plann . inference _
* 19 * 269282 . | we adapt the techniques in stigler [ _ ann .
statist . _
* 1 * ( 1973 ) 472477 ] to obtain a new , general asymptotic result for trimmed @xmath0-statistics via the generalized @xmath1-statistic representation introduced by serfling [ _ ann .
statist . _
* 12 * ( 1984 ) 7686 ] . unlike existing results
, we do not require continuity of an associated distribution at the truncation points .
our results are quite general and are expressed in terms of the quantile function associated with the distribution of the @xmath0-statistic summands .
this approach leads to improved conditions for the asymptotic normality of these trimmed @xmath0-statistics . |
the standard model ( sm ) has been well established with the observation of the last particle predicted in the sm , i.e. , higgs particle , at the lhc experiment@xcite .
the neutrino oscillations with massive neutrinos@xcite provide a strong evidence and a useful window for exploring new physics beyond the sm . in comparison with the quark masses and ckm quark mixing matrix@xcite in the sm ,
the smallness of neutrino masses and large mnsp lepton mixing@xcite have been a long - term puzzle to be understood as a possible indication for new physics .
the greatest success of the sm is the gauge symmetry structure @xmath11 which characterizes three basic forces of strong and electroweak interactions .
all the gauge symmetries are associated with the quantum numbers of quarks and leptons .
@xmath12 characterizes the symmetry among three color quantum numbers of quarks , @xmath13 describes the symmetry between two isospin quantum numbers of quarks and leptons for each family , and @xmath14 is the symmetry corresponding to the hypercharge quantum number of quarks and leptons .
the quark and lepton mixing matrices and cp violations reflect the properties of three family quarks and leptons . to understand the quark and lepton mixing matrices and cp violations
, it is interesting to investigate the possible gauge symmetries among three family quantum numbers .
obviously , a non - abelian gauge family symmetry@xcite for three families of quarks and leptons becomes natural as a simple extension of the sm gauge symmetry structure .
it has been shown in ref .
@xcite that the su@xmath0(3 ) gauge family symmetry enables us to construct a simple gauge family model for understanding the mixing and masses of leptons .
the su@xmath0(3 ) gauge family symmetry was first introduced in early time for estimating the top quark mass@xcite .
it was found in ref .
@xcite that the model can provide a consistent prediction for the lepton mixing and neutrino masses when considering the appropriate vacuum structure of su@xmath0(3 ) gauge symmetry breaking .
specifically , through appropriately making the su@xmath0(3 ) gauge fixing condition with keeping a residual @xmath15-permutation symmetry in the neutrino sector , we can obtain in the neutrino sector the so - called tri - bimaximal mixing matrix@xcite and largely degenerate neutrino masses , while the small mixing matrix in the charged - lepton sector is resulted by requiring the vacuum structure of spontaneous symmetry breaking to possess approximate global u(1 ) family symmetries@xcite .
thus the deviation from tri - bimaximal mixing in the lepton mixing matrix is attributed to the small mixing in the charged - lepton sector , its smallness is protected by the mechanism of approximate global u(1 ) family symmetries @xcite . as the spontaneously symmetry breaking cp - violating phases in the vacuum@xcite
are not restricted by the considered symmetries , they can in principle be large and maximal .
the small masses of the neutrinos and charged leptons are simply ascribed to the usual seesaw mechanism . as a simple case ,
when applying the wolfenstein parametrization@xcite for the ckm quark mixing matrix to the charged - lepton mixing matrix with a similar hierarchy structure as the ckm quark mixing matrix , and making a naive ansatz that all the smallness due to the approximate global u(1 ) family symmetries is characterized by a single wolfenstein parameter @xmath16 , we can obtain an interesting prediction for the lepton mixing matrices with a maximal spontaneous cp violation @xmath17@xcite @xmath18 which agrees with the current experimental data@xcite .
the corresponding leptonic jarlskog cp - violating invariant quantity@xcite reaches the maximal value @xmath19 the resulting neutrino masses are largely degenerate with the value at the order @xmath20 ev with a total mass @xmath21 ev , which is much larger than the minimal limit @xmath22 ev and is expected to be tested by the future experiments .
it is widely expected that the leptonic cp violation can be maximal or large enough to account for the observed matter - antimatter asymmetry in the universe via the leptogenesis mechanism as the cp violation in the sm is not enough to understand the baryogenesis . in this note
, we are going to make a general analysis on the leptonic cp - violating phase and its correlation with the deviation from the tri - bimaximal neutrino mixing based on the current experimental results .
let us begin with the following general structure of mnsp lepton mixing matrix @xmath23 which has been expressed into the standard form with @xmath24@xcite .
an interesting symmetric parametrization was discussed in@xcite . where @xmath25 is a diagonal matrix of phase factors and
can be rotated away by the redefinition of charged lepton fields , and @xmath26 are the so - called majorana phases for majorana neutrinos . it is known that the lepton mixing matrix generally arises from two mixing matrices @xmath27 and @xmath28 , they correspond to the charged - lepton and neutrino mixing matrices arising from diagonalizing the charged - lepton mass matrix and neutrino mass matrix respectively .
when the charged - lepton mass matrix is hermitian @xmath29 with @xmath30 the diagonal mass matrix of charged leptons @xmath31 , the unitary charged - lepton mixing matrix @xmath27 can in general be written as @xmath32 where @xmath33 is a unitary matrix with cp - violating phase @xmath34 , and mixing angles @xmath35 .
@xmath36 is a diagonal phase matrix with three phases @xmath37 , while only two relative phases @xmath38 are physically observable cp - violating phases . for the neutrino mixing matrix , when an appropriate @xmath15-symmetric neutrino mass matrix between the second and third neutrinos is considered to have three independent matrix elements , the resulting neutrino mixing matrix is completely determined to be @xmath39 which is the so - called tri - bimaximal neutrino mixing matrix .
it is noticed that when three phases @xmath40 ( @xmath41 ) in the hermitian mass matrix are not independent and they are related via @xmath42 , one then has @xmath43 on the other hand , when @xmath44 is small in comparison with @xmath45 and @xmath46 , i.e. , @xmath47 , it is easily seen that @xmath34 will not be a dominant source of leptonic cp violation as the cp - violating phase @xmath34 is associated with the mixing angle @xmath44 . in this situation
, we may neglect the effect of cp - violating phase @xmath34 and take a typical case @xmath48 for simplicity of discussions . with these considerations
, we may replace the unitary matrix @xmath33 by an orthogonal rotation matrix @xmath49 @xmath50 to investigate whether the leptonic cp violation can be maximally large with the present experimental measurements on the three mixing angles and the lepton mixing matrix can be characterized by the wolfenstein parametrization method , we may make a sensible analysis by simply taking @xmath51 . in ref.@xcite , the angle @xmath44 is assumed to be zero , thus the effect of @xmath34 automatically disappears .
an alternative consideration was analyzed in@xcite , where the phase @xmath34 was assumed to be the only cp - violating source and the phases @xmath37 are taken to be zero , i.e. , @xmath52 .
the leptonic mixing angles @xmath53 and mass - square differences @xmath54 have been measured by many experiments including the solar neutrino experiment , atmospheric neutrino experiment , accelerator experiment and reactor experiment .
the best - fit results presented in pdg@xcite are @xmath55 which slightly deviates from the tri - bimaximal neutrino mixing .
note that the presently extracted mixing angles from experiments are not sensitive to the cp - violating phase due to the smallness of the effects concerning the cp violation . as the leptonic cp violation is strongly correlated to the non - zero @xmath56 which characterizes the deviation from the tri - bimaximal neutrino mixing matrix
, it is then interesting to investigate the leptonic cp - violating phase and its correlation with the deviation from the tri - bimaximal neutrino mixing based on the above structure of lepton mixing matrix and the current experimental results .
it is seen that the deviation from tri - bimaximal neutrino mixing is described by the charged - lepton mixing matrix @xmath33 or orthogonal rotation matrix @xmath57 , the smallness of the mixing angle @xmath56 indicates that @xmath58 , which motivates us to parameterize the rotation matrix @xmath57 via the wolfenstein parametrization@xcite with a hierarchy structure similar to the ckm quark mixing matrix . with the leptonic wolfenstein parameter @xmath59
, the charged - lepton mixing matrix can be written , to the order @xmath60 , as the following form : @xmath61 where the phase matrix @xmath62 can be absorbed by the redefinitions of charged lepton fields .
note that there is no corresponding wolfenstein parameter @xmath63 in the above parametrization as we have neglected the cp - violating phase @xmath34 .
thus the lepton mixing matrix is given by @xmath64 which shows that based on the tri - bimaximal neutrino mixing , the lepton mixing matrix can be parametrized by three leptonic wolfenstein parameters : @xmath65 , and cp - violating phases @xmath66 ( @xmath67 ) with @xmath68 . as indicated from @xmath15 symmetry of vacuum structure in the @xmath69 model@xcite
, it is reasonable to assume that @xmath70 .
when expressing the lepton mixing matrix @xmath71 to be the standard form by requiring the matrix elements @xmath72 be real with keeping two independent majorana phases , we can read off the leptonic cp - violating phase from @xmath73 @xmath74 the wolfenstein parametrization of lepton mixing matrix is simplified to be @xmath75 in terms of the standard form eq.([eq : st_form ] ) , the lepton mixing matrix can be rewritten in terms of the leptonic wolfenstein parameters as the following form @xmath76 where @xmath77 are the matrix elements of @xmath71 via eq.([eq : msnp_2 ] ) .
the two majorana phases @xmath78 turn out to be @xmath79 and the mixing angle @xmath53 can be expressed in terms of the leptonic wolfenstein parameters as @xmath80 which shows that the leptonic wolfenstein parameters @xmath65 and the cp - violating phase @xmath1 characterize the lepton mixing with deviation from the tri - bimaximal mixing . as an illustration , it is interesting to observe that by taking the leptonic wolfenstein parameters to be the following typical values with a maximal cp - violating phase @xmath81 we obtain the predictions for the lepton mixing angles @xmath82 which are consistent with the pdg s best - fit results given in eq .
( [ eq : para_pdg ] ) @xcite at @xmath83 level , except a small mismatch of @xmath84 .
such a consistency shows that the leptonic wolfenstein parameters chosen in eq.([eq : parameter ] ) are in the reasonable region of parameter space .
alternatively , we may use the pdg s value of @xmath84 given in eq.([eq : para_pdg ] ) @xmath85 to extract the leptonic cp - violating phase @xmath86 . with other parameters chosen as eq.([eq : parameter ] ) , it is easily found that @xmath87 which is very close to the maximal cp - violating phase @xmath88 .
the corresponding two majorana phases with the input parameters as eq.([eq : parameter ] ) are yielded to be @xmath89 we shall make a general constraint on the leptonic wolfenstein parameters @xmath65 and the cp - violating phase @xmath1 by a detailed analysis below .
as it is shown in previous section that the deviation from tri - bimaximal lepton mixing matrix can be described by three leptonic wolfenstein parameters @xmath90 . it is seen from eq.([eq : relation_s13 ] ) and eq.([eq : relation_s23 ] ) that @xmath91 depends on @xmath2 and @xmath92 , while @xmath93 relies on @xmath2 and @xmath3 . in this section
, we shall take the mixing angles @xmath56 and @xmath94 indicated from the measurements as the input to provide a general constraint on leptonic wolfenstein parameters @xmath90 .
the precise measurements on @xmath56 have been carried out by dayabay collaboration group@xcite and reno collaboration group@xcite .
these two experiments measured the disappearance of @xmath95 from the reactor .
the @xmath96 dominated amplitude is given by @xmath97 which shows that the measured results do not sensitively correlate to the values of other two mixing angles @xmath98 and cp - violating phase @xmath1 .
from eq.([eq : relation_s13 ] ) , it is seen that @xmath56 is also insensitive to the cp - violating phase @xmath1 .
thus we may use the experimental data on @xmath56 to make constraints on the leptonic wolfenstein parameters . before doing that
, it is noticed that when keeping the expansion of lepton mixing matrix to the order @xmath99 , we arrive at the following simple relation @xmath100 where we have used the best fit values @xmath101 and @xmath102@xcite to yield the numerical value @xmath103 , which is very close to the wolfenstein parameter of cabibbo angle @xmath104 in quark sector@xcite .
this observation checks the consistence of the assumption that @xmath105 .
let us now turn to make a general analysis by adopting the precisely measured mixing angle @xmath56@xcite @xmath106 which enables us to constrain the allowed region of the combined leptonic wolfenstein parameters @xmath92 for a given @xmath2 . for @xmath107 as a function of @xmath2 .
the contours show the best fit value and @xmath108 deviations .
, width=240 ] the contour plot for the input @xmath109 is shown in figure[fig : cont_13 ] .
it is seen from figure [ fig : cont_13 ] that @xmath110 only occur for small values of @xmath2 . in the plot ,
we have restricted the region to be in the range @xmath111 , so that it satisfies the perturbative expanding of wolfenstein parametrization .
it also leads to a reasonable region for the parameter @xmath2 @xmath112 which will be taken to be a possible allowed region when considering constraints from other two mixing angles @xmath84 and @xmath94 . from eq.([eq : relation_s13 ] ) , it is seen that @xmath113 is an even function of @xmath114 .
thus for the region with @xmath115 , the contours of @xmath113 are just the mirror images of figure [ fig : cont_13 ] , which is omitted here . for tri - bimaximal mixing ,
there is a maximal mixing @xmath116 .
the small deviation to the maximal mixing indicates that the leptonic wolfenstein parameters should be small , which may still be consistent with the current data within the experimental errors . while the recent global fitting results appear to indicate a quite large deviation from the maximal mixing with @xmath117 @xcite , and @xmath118 @xcite . for a general discussion
, we may consider a constraint from a wide range of @xmath94 by covering over different global fitting results , i.e. , @xmath119 .
the resulting constraint is shown in the left panel of figure [ fig : cont_23 ] for parameter @xmath3 as a function of @xmath2 .
for @xmath3 as a function of @xmath2 .
the right panel is the contour plots of @xmath120 for @xmath4 as a function of @xmath2 , where the central value of @xmath121 has been used to fix @xmath92 for a given @xmath2 .
the vertical line labels a critical value of @xmath122 , where @xmath123.,title="fig:",width=240 ] for @xmath3 as a function of @xmath2 .
the right panel is the contour plots of @xmath120 for @xmath4 as a function of @xmath2 , where the central value of @xmath121 has been used to fix @xmath92 for a given @xmath2 .
the vertical line labels a critical value of @xmath122 , where @xmath123.,title="fig:",width=244 ] by combining the constraint from @xmath56 and @xmath94 , we are able to obtain the constraint for the allowed region of @xmath4 as a function of @xmath2 . as shown in the right panel of figure[fig : cont_23 ] , by taking the central value of @xmath121 given in eq.([eq : daya_reno_13 ] ) , we can obtain the allowed region for @xmath4 as a function of @xmath114 from the given values of @xmath120 .
it is seen that a wide region swept by the curve when @xmath120 increasing from @xmath124 to @xmath125 is allowed .
note that there is a special situation that for @xmath126 , then @xmath127 , namely @xmath128 for @xmath129 . as a consequence ,
four curves intersect with each other at this point , as indicated in figure[fig : cont_23 ] .
in this section , we should make a general analysis on the leptonic cp - violating phase and its correlation to the deviation from the tri - bimaximal neutrino mixing , which is characterized by the leptonic wolfenstein parameters as discussed in the previous section .
it is seen from eq .
( [ eq : relation_s12 ] ) that @xmath130 depends on cp - violating phase @xmath1 , @xmath2 , @xmath92 . here
@xmath92 can be constrained from @xmath56 for a given @xmath2 .
the mixing angle @xmath84 is well determined from solar neutrino oscillation experiments .
the measured value of @xmath84 generally correlates to the value of @xmath94 .
it is convenient to obtain the values of @xmath84 by setting @xmath131 .
the global fitting results have provided us with both values of @xmath94 and @xmath84 .
although a non - maximal @xmath94 is hinted@xcite , there is no tension among different global fitting results on @xmath84 .
for instance , @xmath132@xcite , @xmath133@xcite , and @xmath134@xcite . here
we take the result @xmath135 given in pdg@xcite to make constraints on the cp - violating phase @xmath1 as a function of @xmath2 .
the value of @xmath92 is constrained from @xmath56 .
the allowed region for cp - violating phase @xmath1 is given as a function of @xmath2 in figure[fig : cont_12 ] , where we have taken the central value @xmath136 to yield the value of @xmath92 .
for the cp - violating phase @xmath1 as a function of @xmath2 with the best fit value of @xmath137 and @xmath138 deviations . for @xmath139 ,
the allowed @xmath86 ranges from @xmath140 to @xmath141 , and for @xmath142 , the resulting @xmath86 is close to maximal @xmath141.,width=240 ] it is seen from figure [ fig : cont_12 ] that there are two special regions : for @xmath139 , the values of @xmath143 is insensitive to @xmath144 , the allowed region of @xmath1 ranges from @xmath145 to @xmath141 . while for @xmath146
, the constraint on @xmath144 becomes very strong , the resulting cp - violating phase is near maximal @xmath147 . in this region
, @xmath148 is insensitive to the values of @xmath114 .
thus the leptonic cp violation favors a maximal cp violation for a large range of leptonic wolfenstein parameter @xmath2 .
note that a minimal cp - violating phase @xmath149 was obtained in a global fit @xcite when the atmospheric neutrino data are included , while such a fitting result corresponds to a special region in the parameter space , which does not exclude a large or nearly maximal cp violation .
it is useful to combine all the constraints obtained from @xmath121 , @xmath120 and @xmath148 and plot them together in the same figure[fig : final ] , so that it is easily seen the allowed values of @xmath3 , @xmath4 and @xmath1 for a given @xmath114 .
a big uncertainty arises from whether @xmath120 is largely deviate from the maximal mixing , which makes the allowed values of @xmath3 and @xmath4 become large . it is easy to see from figure[fig : final ] that there are two typical regions for leptonic wolfenstein parameters characterized with a small @xmath2 ( @xmath150 ) and a large @xmath2 ( @xmath151 ) . for the small @xmath152
, we have @xmath153,\ ; a_e\in[12 , 2],\ ; \rho_e\in[-12,-2],\;\ ; |{\delta}_{e}|\in[\pi , \pi/2 ) .
\end{aligned}\ ] ] which shows that the cp - violating phase @xmath144 is not well constrained in this case .
a global fitting result cited in@xcite corresponds to a solution of the small @xmath114 with @xmath154 . from the results given in@xcite for the normal hierarchy : @xmath117 , @xmath155 , @xmath156 , and @xmath157 , one can easily read from fig.[fig : final ] the corresponding leptonic wolfenstein parameters @xmath158 with the central values , two majorana phases are found to be very small @xmath159 for the case of inverted hierarchy , the result is very close to the above one , we shall omit it here . for large values of @xmath2 , we have @xmath160,\ ; a_e\in[7 , 0],\;\ ; \rho_e\in[-10,15 ] , \ ; \ ; |{\delta}_{e}|\sim [ 3\pi/4 , \pi/2 ) . \end{aligned}\ ] ] where the cp - violating phase @xmath144 is strongly constrained , only a large or nearly maximal cp violation is favorable .
it is interesting to observe from fig .
[ fig : final ] that when taking the value of the leptonic wolfenstein parameter @xmath2 to be the same as the one in the quark sector , @xmath161 , and fixing the lepton mixing angles to be the central values @xmath162 and @xmath163 , we arrive at a sensible result for the leptonic wolfenstein parameters @xmath164 which is compatible with the wolfenstein parameters in quark sector @xmath165 in this case , the resulting lepton mixing matrix is given by @xmath166 to show manifestly such an interesting situation , it is useful to plot the leptonic wolfenstein parameters in a parameter space as shown in fig.[fig : para_space ]
. it is easily seen that when fixing the parameter @xmath167 , the whole parameters space for @xmath168 and @xmath86 is almost located on two planes with @xmath169 .
the above analysis implies that only a large or nearly maximal leptonic cp violation is favorable in a large region of parameter space when @xmath170 .
the above results indicate a strong correlation between charged - leptons and quarks .
an assumption that @xmath171 and @xmath172 was discussed early in@xcite .
we have shown that the lepton mixing can be parametrized by the wolfenstein parametrization method based on a general structure of lepton mixing matrix , where the mixing matrix from neutrino sector is a tri - bimaximal mixing and the mixing matrix from charged - lepton has small mixing .
such a structure of lepton mixing has been shown to be resulted from the su@xmath0(3 ) gauge family model@xcite when considering the appropriate vacuum structure of su@xmath0(3 ) gauge symmetry breaking . where the tri - bimaximal mixing can be yielded from the residual @xmath15-permutation symmetry in the neutrino sector and the small mixing in the charged - lepton sector is led by requiring the vacuum structure of spontaneous symmetry breaking to possess approximate global u(1 ) family symmetries .
we have demonstrated that the small mixing matrix in the charged - lepton sector characterizes the deviation from tri - bimaximal mixing in the lepton mixing matrix , and can be parametrized by the wolfenstein parametrization method .
as the spontaneous cp - violating phases in the vacuum are in general not restricted by the considered symmetries , so that they can in principle be large and maximal .
based on the input values of lepton mixing angles @xmath56 , @xmath94 and @xmath84 indicated from various neutrino experiments , we have made a general analysis for the allowed leptonic cp - violating phase @xmath1 and leptonic wolfenstein parameters @xmath2 , @xmath3 , @xmath4 .
it has explicitly been shown how the leptonic cp violation correlates to the leptonic wolfenstein parameters which characterize the deviation of tri - bimaximal lepton mixing . for a reasonable range of parameter @xmath173
, there appear two typical regions , i.e. , one with @xmath174 , and other with @xmath175 .
for the small values of @xmath174 , the mixing angles @xmath176 are insensitive to @xmath144 , thus the cp - violating phase @xmath1 is not well constrained , its allowed region can range from @xmath177 to @xmath178 . while for the large values of @xmath175 , the cp - violating phase @xmath1 has strongly been constrained , only a large or nearly maximal leptonic cp violation with @xmath179 is allowed .
it has been demonstrated that when taking the leptonic wolfenstein parameter @xmath2 to be the cabbibo angle in quark sector , @xmath7 , we are able to obtain a sensible result with @xmath180 , @xmath8 , @xmath9 , @xmath10 , which is compatible with the wolfenstein parameters in quark sector : @xmath181 , @xmath182 , @xmath183 , @xmath184 .
such a correlation implies a possible common origin of masses and mixing angles for the charged - leptons and quarks . in conclusion , the lepton mixing matrix
can well be characterized by leptonic wolfenstein parameters in the basis of tri - bimaximal neutrino mixing .
the leptonic cp violation has a strong correlation to the leptonic wolfenstein parameters , a large or nearly maximal leptonic cp violation is favorable in a large region of parameters .
more precise measurements for the lepton mixing angles are very helpful .
it is essential to have a direct measurement for the leptonic cp violation in near future .
the authors would like to thank petcov and valle for useful discussions and comments .
this work is supported in part by the national nature science foundation of china ( nsfc ) under grants no .
10975170 , no.10905084,no.10821504 ; and the project of knowledge innovation program ( pkip ) of the chinese academy of science .
99 g. aad et al .
[ atlas collaboration ] , phys .
b * 716*.1 ( 2012 ) . s. chatrchyan et al .
[ cms collaboration ] , phys .
b * 716*.30 ( 2012 ) . y. fukuda et al .
[ super - kamiokande collaboration ] , phys .
81 , 1562 ( 1998 ) .
the sno collaboration , phys .
87 , 71301 ( 2001 ) .
k. eguchi et al .
[ kamland collaboration ] , phys .
90 , 021801 ( 2003 ) ; phys .
lett . * 94 * , 081801(2005 ) .
m. sanchez et al .
( soudan 2 collaboration ) , phys . rev . *
d68 * , 113004(2003 ) .
m. ambrosio et al .
( marco collaboration ) , eur .
j. * c36 * , 323(2004 ) .
e. aliu et al .
[ k2k collaboration ] , phys .
94 , 081802 ( 2005 ) .
y. ashie et al .
[ sk collaboration ] , phys .
d71 ( 2005 ) .
m. altmann et al .
[ gno collaboration ] , phys .
b616 , 174 ( 2005 ) .
m. apolonio et al .
[ chooz collaboration ] , eur .
j. c27 , 331 ( 2003 ) .
k. abe et al .
[ t2k collaboration ] , phys .
107 , 041801 ( 2011 ) , arxiv:1106.2822 [ hep - ex ] .
p. adamson et al .
[ minos collaboration ] , phys .
107 , 181802 ( 2011 ) , arxiv:1108.0015 [ hep - ex ] .
y. l. wu , phys .
d * 77 * , 113009 ( 2008 ) [ arxiv:0708.0867 [ hep - ph ] ] .
y. l. wu , int .
j. mod .
a * 23 * , 3376 ( 2008 ) [ arxiv:0807.3847 [ hep - ph ] ] .
y. l. wu , phys.rev .
* d60 * ( 1999 ) 073010 . y. l. wu , nucl.phys.proc.suppl . * 85 * ( 2000 ) 193 .
y. l. wu , invited talk at the 30th international conference on high - energy physics ( ichep 2000 ) , osaka , japan . y. l. wu , eur.phys.j .
* c10 * ( 1999 ) 491 .
y. l. wu , j. phys .
g : nucl . part .
* 26 * 1131 ( 2000 ) .
y. l. wu , science in china * a43 * ( 2000 ) 988 .
c. carone and m. sher , phys
b420 * , 83 ( 1998 ) .
e. ma , phys.lett .
b456 , 48 ( 1999 ) , hep - ph/9812344 . c. wetterich , phys.lett .
b451 , 397 ( 1999 ) , hep - ph/9812426 .
r. barbieri , l.j .
hall , g.l .
kane and g.g .
ross , hep - ph/9901228 .
t. yanagida , phys .
d * 20 * , 2986 ( 1979 ) .
p. f. harrison , d. h. perkins and w. g. scott , phys .
* b 530 * , 167 ( 2002 ) . z .-
xing , phys . lett .
* b533 * , 85(2002 ) .
p. f. harrison and w.g .
scott , phys .
b535 * , 163(2002 ) .
harrison and w.g .
scott , phys .
b557 * , 76(2003 ) .
x. g. he and a. zee , phys .
b560 * , 87(2003 ) .
l.j . hall and s.
weinberg , phys .
* d48 * , 979 ( 1993 ) . y. l. wu and l. wolfenstein , phys .
rev . lett .
* 73 * , 1762 ( 1994 ) [ arxiv : hep - ph/9409421 ]
. l. wolfenstein and y. l. wu , phys .
lett . * 73 * , 2809 ( 1994 ) [ arxiv : hep - ph/9410253 ] .
y. l. wu , arxiv : hep - ph/9404241 ; see also : invited talk at 5th conference on the intersections of particle and nuclear physics , st .
petersburg , fl , 31 may- 6 jun 1994 , published in proceedings , pp338 , edited by s.j .
seestrom ( aip , new york , 1995 ) , hep - ph/9406306 . c. jarlskog , phys .
55 , 1039 ( 1985 ) ; d.d .
wu , phys .
d 33 , 860 ( 1986 ) .
j. schechter and j.w .
f. valle , phys .
d 22 , 2227 ( 1980 ) ; + h. nunokawa , s. parke , j. w.f .
valle , progress in part . and nucl .
60 , 338 ( 2008 ) ; + w. rodejohann and j. w. f. valle , phys .
d84 , 073011 ( 2012 ) . | we investigate a general structure of lepton mixing matrix resulting from the su@xmath0(3 ) gauge family model with an appropriate vacuum structure of su@xmath0(3 ) symmetry breaking .
it is shown that the lepton mixing matrix can be parametrized by using the wolfenstein parametrization method to characterize its deviation from the tri - bimaximal mixing .
a general analysis for the allowed leptonic cp - violating phase @xmath1 and the leptonic wolfenstein parameters @xmath2 , @xmath3 , @xmath4 is carried out based on the observed lepton mixing angles .
we demonstrate how the leptonic cp violation correlates to the leptonic wolfenstein parameters .
it is found that the phase @xmath1 is strongly constrained and only a large or nearly maximal leptonic cp - violating phase @xmath5 is favorable when @xmath6 . in particular , when taking @xmath2 to be the cabbibo angle @xmath7 , a sensible result for leptonic wolfenstein parameters and cp violation is obtained with @xmath8 , @xmath9 , @xmath10 , which is compatible with the one in quark sector .
an interesting correlation between leptons and quarks is observed , which indicates a possible common origin of masses and mixing for the charged - leptons and quarks . |
a topological superconductor has topologically protected gapless edge states , some of which can host majorana fermions @xcite .
there are two kinds of edge states in the topological superconductor .
one is a dirac cone , the other is a flat band , namely , dispersionless zero - energy state @xcite
. the dirac cone can be found in the fully gapped topological superconductors when the chern number of the occupied bands is nonzero . however , the flat band can appear in the fully gapped topological superconductors as well as the gapless topological superconductors which , apart from the particle - hole symmetry , have some extra symmetries in the hamiltonian .
such flat bands are known to occur at the zigzag and bearded edge in graphene @xcite , in the noncentrosymmetric superconductor @xcite and in other systems with topologically stable dirac points @xcite .
the flat dispersion implies a peak in the density of states which has a visible experimental signature in the tunneling conductance measurements @xcite .
the zero - bias conductance peak has been observed in recent experiments on the insb nanowire @xcite and @xmath3 @xcite and might be due to the flat bands . in this paper , we shall investigate the spin - singlet superconductor with the rashba and dresselhaus ( 110 ) spin - orbit couplings .
especially , we focus on the hamiltonian with spin - orbit coupling of dresselhaus ( 110 ) type which is a gapless topological system containing two kinds of edge states mentioned above .
we also apply a magnetic field to the superconductor and consider the zeeman effect .
the bogoliubov - de gennes ( bdg ) hamiltonian of the superconductor is generically particle - hole symmetric so that we can associate a pfaffian invariant @xmath0 with it as a topological invariant of the system .
the pfaffian invariant @xmath0 can be used in distinguishing the topologically nontrivial phase from the trivial one .
the nontrivial topological phase in this bdg hamiltonian is a majorana type which can be exploited for implementing fault - tolerant topological quantum computing schemes that are inherently decoherence - free @xcite .
all the possible phase diagrams in the bdg hamiltonian are demonstrated in our study .
furthermore , we find that the bdg hamiltonian can have partial particle - hole symmetry and chiral symmetry which can be used to define the one dimensional pfaffian invariant @xmath1 and the winding number @xmath2
. the pfaffian invariant @xmath1 and the winding number @xmath2 can be used in determining the location of the zero - energy majorana flat band .
we show that only when the system has these extra symmetries , the majorana flat band will emerge so that a zero - bias conductance peak in the tunneling conductance measurements can be observed @xcite .
the paper is organized as follows .
the bdg hamiltonian for the spin - singlet superconductor with the rashba and dresselhaus ( 110 ) spin - orbit couplings is given in sec .
[ model ] . in sec .
[ symofbdg ] , we discuss the symmetries of the bdg hamiltonian and the topological invariants associated with these symmetries are given in sec .
[ tiofbdg ] .
all the possible phase diagrams of the bdg hamiltonian are discussed in sec .
[ pdofbdg ] whilst the edge spectra of the bdg hamiltonian are demonstrated in sec .
[ mbsofbdg ] . finally , we give a brief summary in sec .
[ summary ] .
we model the spin - singlet superconductor on a square lattice .
the kinetic energy is @xmath4 where @xmath5 is the creation ( annihilation ) operator of the electron with spin @xmath6 at site @xmath7 , @xmath8 @xmath9 is the unit vector in the @xmath10 ( @xmath11 ) direction , and @xmath12 is the hopping amplitude and @xmath13 is the chemical potential . for the spin - singlet superconductor
, we study the @xmath14-wave and @xmath15-wave pairings in this work .
the @xmath14-wave superconducting term in the square lattice is @xmath16 .
\end{split}\ ] ] similarly , the @xmath15-wave superconducting term is @xmath17.\\ \end{split}\ ] ] we assume that all the superconducting parameters @xmath18 , @xmath19 , @xmath20 and @xmath21 are uniform in the whole superconductor
. the spin - orbit couplings can arise from structure inversion asymmetry of a confinement potential ( e.g. , external electric field ) or bulk inversion asymmetry of the underlying crystal ( e.g. , the zinc blende structure ) @xcite .
these two kinds of asymmetries lead to the well - known rashba and dresselhaus spin - orbit couplings .
the rashba spin - orbit coupling in the square lattice is of the form @xmath22,\\ \end{split}\ ] ] where @xmath23 is the coupling strength of the rashba spin - orbit coupling .
the dresselhaus ( 110 ) spin - orbit coupling is formulated as @xmath24 where @xmath25 is the coupling strength for the dresselhaus ( 110 ) spin - orbit coupling .
( 110 ) are the common - used miller index .
we also apply an arbitrary magnetic field to the superconductor . by neglecting the orbital effect of the magnetic field
, we consider the zeeman effect as @xmath26 where @xmath27 and @xmath28 are pauli matrices operating on spin space .
therefore , the spin - singlet superconductor with the rashba and dresselhaus ( 110 ) spin - orbit couplings in an arbitrary magnetic field is dictated by the hamiltonian @xmath29 .
in the momentum space , the hamiltonian is recast into @xmath30 with @xmath31 , where @xmath32 . finally , the bogoliubov - de gennes hamiltonian @xmath33 for the superconductor is @xmath34 \end{split}\ ] ] where @xmath35 , @xmath36 and @xmath37 .
for the general bdg hamiltonian eq .
( [ hbdg ] ) , it satisfies the particle - hole symmetry @xmath38 where @xmath39 , @xmath40 and @xmath41 is the complex conjugation operator .
we find that apart from the particle - hole symmetry , the bdg hamiltonian can satisfy some extra symmetries , namely , partial particle - hole symmetry , chiral symmetry and partial chiral symmetry when we set some parameters in the hamiltonian eq .
( [ hbdg ] ) to @xmath42 .
the particle - hole-@xmath43 and particle - hole-@xmath44 symmetries are defined as @xmath45 and @xmath46 where @xmath47 ( @xmath48 ) takes the @xmath43 ( @xmath44 ) in the hamiltonian to @xmath49 ( @xmath50 ) .
the chiral symmetry is @xmath51 the chiral-@xmath43 and chiral-@xmath44 symmetries are defined as @xmath52 and @xmath53 where @xmath54 ( @xmath55 ) takes the @xmath43 ( @xmath44 ) in the hamiltonian to @xmath49 ( @xmath50 ) .
we are interested in the bdg hamiltonian which has one or more extra symmetries . in the following sections
, we would like to consider these kinds of bdg hamiltonian as listed in tab .
( [ symmandti ] ) .
the spin - singlet superconductor with rashba spin - orbit coupling has been extensively investigated in the reference @xcite .
here we only consider the general @xmath56 pairing case ( a ) for the superconductor with rashba spin - orbit coupling .
we shall focus on the superconductor with dresselhaus ( 110 ) spin - orbit coupling as shown in the case ( b)-(g ) in tab .
( [ symmandti ] ) . [ cols="^,^,^,^,^,^",options="header " , ] -wave superconductor with dresselhaus ( 110 ) spin - orbit coupling in the edge brillouin zone of @xmath57 .
@xmath58 is the lattice site .
@xmath59 is the probability of quasiparticle at site @xmath58.,width=302 ]
in summary , we have investigated the topological phase and the majorana bound state in the spin - singlet superconductor with the rashba and dresselhaus ( 110 ) spin - orbit couplings .
we find that apart from the particle - hole symmetry , the bdg hamiltonian can also exhibit the chiral symmetry , partial particle - hole symmetry and partial chiral symmetry as listed in tab .
( [ symmandti ] ) .
the topological invariants corresponding to these symmetries have been discussed in sec .
( [ tiofbdg ] ) and are also listed in tab .
( [ symmandti ] ) .
we have demonstrated all the possible phase diagrams of the spin - singlet superconductor with the rashba and dresselhaus ( 110 ) spin - orbit couplings in sec .
( [ pdofbdg ] ) by the pfaffian invariant @xmath0 of the particle - hole symmetric hamiltonian .
we find that only when the hamiltonian has partial particle - hole symmetry or chiral symmetry , the edge spectrum is flat band protected by the one dimensional pfaffian invariant @xmath1 or the winding number @xmath2 ; otherwise the edge spectrum is dirac cone .
the zero - energy state of the dirac cones and the zero - energy flat bands are the majorana type which is a precious source for topological quantum computing .
the edge spectra of the cases listed in tab .
( [ symmandti ] ) are shown in sec .
( [ mbsofbdg ] ) .
we find that the pfaffian invariant @xmath1 and the winding number @xmath2 can be used in determining the location of the zero - energy flat bands .
26ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1103/physrevb.61.10267 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.100.096407 [ * * , ( ) ] link:\doibase 10.1103/physrevb.82.134521 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.125318 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.177002 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.108.107005 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.150408 [ * * , ( ) ]
link:\doibase 10.1103/physrevb.87.054501 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.104.040502 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.217001 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.060504 [ * * , ( ) ] link:\doibase 10.1103/physrevb.86.161108 [ * * , ( ) ] link:\doibase 10.1103/physrevb.83.224511 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.020501 [ * * , ( ) ] @noop link:\doibase 10.1103/physrevb.54.17954 [ * * , ( ) ] link:\doibase 10.1134/s0021364011020147 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.097002 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevlett.107.217001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.108.057001 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/revmodphys.80.1083 [ * * , ( ) ] @noop _ _ ( , ) link:\doibase 10.1103/physrevb.82.184525 [ * * , ( ) ] | we examine the topological properties of a spin - singlet superconductor with rashba and dresselhaus ( 110 ) spin - orbit couplings .
we demonstrate that there are several topological invariants in the bogoliubov - de gennes ( bdg ) hamiltonian by symmetry analysis .
we use the pfaffian invariant @xmath0 for the particle - hole symmetry to demonstrate all the possible phase diagrams of the bdg hamiltonian .
we find that the edge spectrum is either dirac cone or flat band which supports the emergence of the majorana fermion in this system .
for the majorana flat bands , a higher symmetric bdg hamiltonian is needed to make them topologically stable .
the pfaffian invariant @xmath1 and the winding number @xmath2 are used in determining the location of the majorana flat bands . |
any single - mode gaussian channel @xmath33 can be reduced to a simpler ` canonical form ' up to gaussian unitaries , applied at the input and the output of the channel .
the canonical forms are classified as @xmath137 , @xmath138 , @xmath139 , @xmath140 , @xmath141 and @xmath142 ( see ref .
@xcite for this formalism ) .
they transform the cm @xmath143 of an input state as @xmath144 , with channel matrices @xmath145 and @xmath146 which are diagonal .
the most typical and important canonical forms are phase - insensitive , with @xmath145 and @xmath146 both proportional to the identity , of the form @xmath147 where @xmath110 is the transmissivity and @xmath111 is thermal noise .
the first example is the lossy channel ( form @xmath141 ) for which @xmath47 $ ] and @xmath148 with @xmath149 . in terms of the quadrature operators
@xmath150 its action is expressed by the input - output transformations@xmath151 where @xmath152 are the quadratures of an environmental mode in a thermal state with variance @xmath153 , with @xmath154 being the mean number of thermal photons . at the border , for @xmath155 , this channel coincides with the form @xmath137 which is a completely depolarizing channel . for @xmath49
, it just becomes an identity channel ( a particular instance of the form @xmath140 ) .
the second example is the additive - noise channel ( form @xmath140 ) , which has @xmath49 and @xmath50 .
this channel corresponds to the transformation @xmath156 where @xmath157 is a classical random variable with cm @xmath158 .
the final example of phase - insensitive canonical form is the amplifier channel ( another kind of form @xmath141 ) which has @xmath159 and @xmath160 .
this is realized by @xmath161 where @xmath152 are the quadratures of a thermal mode .
then , we have canonical forms which are sensitive to phase .
the most important is the form @xmath142 which is the conjugate of the amplifier channel , i.e. , the environmental output of a two - mode squeezer which dilates the amplifier channel .
this form can be associated with negative transmissivities @xmath96 and is described by the matrices @xmath162 with @xmath163 and @xmath164 .
its sensitivity to phase is clear from the fact that @xmath165 .
it corresponds to the transformation@xmath166 in the main paper , our phase - insensitive gaussian channel is given by the forms @xmath141 and @xmath140 , while our extended gaussian channel is given by the forms @xmath141 , @xmath140 and @xmath142 .
such an extended channel can be compactly characterized by the matrices@xmath167,~\mathbf{n}=\eta\mathbf{i},\ ] ] with @xmath97 and @xmath46 .
the corresponding input - output relations can be written in the form@xmath168{cc}\sqrt{|\tau| } & 0\\ 0 & [ \mathrm{sign}(\tau)]\sqrt{|\tau|}\end{array } \right ) \mathbf{\hat{x}}+\mathbb{\zeta } , \label{classn}\ ] ] where @xmath169 is a noise - variable with cm equal to @xmath170 .
the output entropy of this extended gaussian channel is minimized by taking a coherent state at the input @xcite .
finally , we have the ` pathological ' canonical forms @xmath138 and @xmath139 , which are highly phase - sensitive .
in particular , form @xmath138 is described by @xmath171 while form @xmath139 is described by @xmath172 form @xmath138 is highly symmetric in the transmission matrix , and represents a sort of ` half ' depolarizing channel which replaces one of the input quadratures with thermal noise .
when this form is locally applied to a bipartite state , it completely destroys the correlations in one of the quadratures .
by contrast , @xmath139 is highly asymmetric in the noise matrix and its effect is to add a unit of vacuum noise to only one of the quadratures . for these pathological forms
, the optimal input state is expected to be a squeezed vacuum .
indeed , this can easily be proven for @xmath139 , for which the optimal input state is an infinitely - squeezed vacuum , with cm @xmath173 where @xmath174 .
in fact , at the output we have a gaussian state with covariance matrix @xmath175 whose determinant goes to @xmath128 , so that it is asympotically pure .
this implies that the output entropy goes to zero , i.e. , to its minimum value .
however , if we enforce an energy constraint at the input of this form , then we do not know the optimal input state which minimizes the output entropy .
the situation is more involved for the other form @xmath138 . when @xmath176 , this is the conjugate channel of @xmath139 in a unitary dilation with a pure environment ( stinespring dilation ) . since the global output of this dilation is pure for an input pure state
( we can always restrict the search to pure states due to the concavity of the entropy ) , we have that the output entropy of the conjugate channel @xmath138 is equal to the output entropy of @xmath139 .
this trivially implies that the infinitely - squeezed vacuum is again the optimal input state .
however , for @xmath177 , the dilation involves a thermal environment , and the optimality of this state is only conjectured for @xmath138 .
furthermore , also for this form , the optimal input is not known if we enforce an energy constraint .
here we prove that a cm in the special normal - form @xmath178 can equivalently be expressed as in eq .
( [ cmab ] ) , i.e. , using the parametrization @xmath179 for some choice of @xmath45 and @xmath54 .
let us start by re - writing the bona - fide conditions for the parameters @xmath41 , @xmath87 , and @xmath108 . from the uncertainty principle
, we derive @xcite @xmath30 and @xmath180 , where@xmath181 for @xmath182 or @xmath183 we must have @xmath184 .
this cm is just realized by taking @xmath155 . for any @xmath185 ,
it is sufficient to prove that we can generate all cms with maximal correlations @xmath186 by means of our parametrization .
if this is possible for maximal correlations , then it will also be possible for intermediate correlations @xmath180 , where conditions are less stringent ( since this extension is merely technical , it has been omitted ) .
first consider the case @xmath187 , so that @xmath188 .
imposing @xmath189 and using eq .
( [ paraapp ] ) , we derive @xmath190 ( possible for @xmath159 ) . using @xmath191 in the parametrization of @xmath41 , we find @xmath192 , which means that we can generate all possible values of @xmath193 by freely varying @xmath50 .
consider now @xmath194 , so that @xmath195 . by imposing @xmath196 and using eq .
( [ paraapp ] ) , we derive @xmath197 , which implies @xmath198 . by replacing @xmath199 in the parametrization of @xmath41
, we derive @xmath200 .
it is clear that we can generate all possible values of @xmath201 by freely varying @xmath202 .
in order to derive the parametrization given in eqs .
( [ a])-([cp ] ) , we work in the heisenberg picture , considering the input - output transformations for the quadrature operators @xmath150 . the action of the first squeezer in fig .
[ decopic ] is given by@xmath203 where @xmath112 is positive and @xmath204{cc}r^{1/2 } & 0\\ 0 & r^{-1/2}\end{array } \right ) .\ ] ] the role of this squeezer is to rectify the input ensemble created by the local gaussian povm @xmath86 with @xmath81 , in such a way to have coherent states on mode @xmath205 .
in fact , applying this local detection to the @xmath6 mode of an epr state @xmath35 with cm@xmath206{cc}b\mathbf{i } & \sqrt{b^{2}-1}\mathbf{c}\\ \sqrt{b^{2}-1}\mathbf{c } & b\mathbf{i}\end{array } \right ) , ~\mathbf{c=\pm}\left ( \begin{array } [ c]{cc}1 & 0\\ 0 & -1 \end{array } \right ) , \ ] ] the other mode @xmath41 is projected onto amplitude - modulated squeezed states with squeezing @xmath89 which ranges between @xmath207 and @xmath87 ( see sec . [ remote_app ] for more details on the remote preparation of gaussian states ) .
after the first squeezer , the action of the extended gaussian channel is expressed by @xmath208{cc}\sqrt{|\tau| } & 0\\ 0 & [ \mathrm{sign}(\tau)]\sqrt{|\tau|}\end{array } \right ) \mathbf{\hat{x}}_{a^{\prime}}+\mathbb{\zeta~},\ ] ] where @xmath169 is a noise - variable with cm@xmath209 , where @xmath54 and @xmath97 ( see eq .
( [ classn ] ) in sec .
[ canforms_app ] ) . finally , the second squeezer realizes@xmath210 where parameter @xmath94 is chosen in such a way to put the output cm @xmath77 in normal - form .
after simple algebra we find @xmath211 using these equations and the fact that mode @xmath6 is subject to the identity ( before detection ) , we have that the input epr state is mapped into the output gaussian state @xmath7 with normal - form cm @xmath212 , where@xmath213\sqrt{\left\vert \tau\right\vert ( b^{2}-1)\frac{\theta(r)}{\theta(r^{-1 } ) } } , \label{cp_app}\ ] ] which correspond to eqs .
( [ a])-([cp ] ) . in order to generate elements of the family
, we randomly pick values for @xmath107 , @xmath103 $ ] , @xmath97 and @xmath54 . then , we compute @xmath41 , @xmath108 , and @xmath109 according to eqs .
( [ a_app])-([cp_app ] ) .
alternatively , we can pick values for @xmath30 and randomly generate values for @xmath112 , @xmath110 , and @xmath111 which are compatible with eq .
( [ a_app ] ) . using these values , we finally compute @xmath108 and @xmath109 according to eqs .
( [ c_app ] ) and ( [ cp_app ] ) .
we adopt this second procedure since it allows us to represent the elements of the family on the correlation plane @xmath114 for given values of @xmath41 and @xmath87 .
fixing a value for @xmath106 , the condition in eq .
( [ a_app ] ) restricts the range of variability of the transmissivity .
in fact , we can express the parameter @xmath111 as function of @xmath41 by inverting eq .
( [ a_app ] ) and finding@xmath214 then , imposing @xmath54 implies @xmath215 $ ] , where@xmath216{c}\frac{b+(br-2)r-\sqrt{(r^{2}-1)^{2}b^{2}-4a^{2}r\gamma_{-}}}{2\gamma_{-}}~~\text{for } a\leq b,\\ \\ \frac{b+(br+2)r+\sqrt{(r^{2}-1)^{2}b^{2}+4a^{2}r\gamma_{+}}}{2\gamma_{+}}~~\text{for } b\leq a , \end{array } \right.\end{aligned}\ ] ] and @xmath217 .
thus , for any values of @xmath30 , we randomly generate the parameters @xmath103 $ ] and @xmath218 $ ] .
then , we compute @xmath111 according to eq .
( [ eta_app ] ) to be replaced in @xmath219 of eq .
( [ theta_app ] ) .
we finally compute the values of @xmath108 and @xmath109 according to eqs .
( [ c_app ] ) and ( [ cp_app ] ) .
let us consider an arbitrary two - mode gaussian state @xmath7 with mean value @xmath220 and cm @xmath27{cc}\mathbf{a } & \mathbf{c}\\ \mathbf{c}^{t } & \mathbf{b}\end{array } \right ) , \label{vabblockform}\ ] ] where @xmath221 , @xmath222 and @xmath223 are @xmath224 real blocks .
mode @xmath6 is detected by an arbitrary gaussian measurement , which is defined as a povm with measurement operator @xmath225 where @xmath226 is the complex outcome , @xmath227 is the displacement operator @xcite with annihilation operator @xmath228 , and @xmath229 is a zero - mean gaussian state with cm @xmath230 .
this operator displaces the state by @xmath231 and projects it onto the gaussian state @xmath229 .
the gaussian measurement is rank - one when @xmath229 is pure .
the simplest choice is the vacuum state @xmath232 , in which case the gaussian measurement corresponds to heterodyne detection .
the gaussian povm can also be expressed in terms of the quadrature operators @xmath233 and the real vector @xmath234 .
in fact , from the decomposition of the identity@xmath235 we derive the equivalent measurement operator@xmath236 where @xmath237 and@xmath238{cc}0 & 1\\ -1 & 0 \end{array } \right ) .
\label{weyl1}\ ] ] by detecting mode @xmath6 , the outcome @xmath239 is achieved with probability @xmath240 } { 2\pi\sqrt{\det(\mathbf{b}+\mathbf{v}_{0})}},~\mathbf{d}:=\mathbf{\bar{x}}_{b}-\mathbf{k } , \label{px_for}\ ] ] which is gaussian with classical cm @xmath241 .
correspondingly , the other mode @xmath5 is projected into a conditional gaussian state @xmath242 with mean value@xmath243 and cm given by @xmath244 if we detect mode @xmath5 with outcome @xmath239 , then we have to permute @xmath5 and @xmath6 in the formulas .
this means that @xmath245 and @xmath246 are replaced by @xmath247 and @xmath248 .
then , eqs .
( [ px_for ] ) , ( [ xcond_for ] ) and ( [ cmcond_for ] ) , are subject to the replacements @xmath249 , @xmath250 , @xmath251 @xmath252 and @xmath253 . despite the fact that eq .
( [ cmcond_for ] ) is well - known in the literature @xcite , the other two eqs .
( [ px_for ] ) and ( [ xcond_for ] ) are rarely - used and almost unknown . for the sake of completeness and pedagogical reasons
we prove all three formulas in sec .
[ appmiscellanea ] . for its relevance to our work
, we consider the specific case where the two - mode gaussian state @xmath7 is an epr state , and mode @xmath6 is heterodyned , so that coherent states are prepared on mode @xmath5 . let us start with one type of epr state , whose cm @xmath77 in eq .
( [ vabblockform ] ) has blocks @xmath254 where @xmath255 and @xmath256 .
afterwards we deal with the other epr type ( @xmath257 ) .
since mode @xmath6 is heterodyned , we have @xmath258 in the previous formulas . according to eq .
( [ px_for ] ) , the outcome @xmath239 follows a gaussian probability @xmath259 with classical cm @xmath260 , as expected since the reduced state @xmath12 is thermal with cm @xmath261 .
correspondingly , the other epr mode @xmath5 is projected into a conditional gaussian state @xmath242 with mean - value @xmath262 from eq .
( [ xcond_for ] ) , and cm @xmath263 , from eq . ( [ cmcond_for ] ) .
it is therefore a coherent state with a gaussianly - modulated mean - value @xmath264 . in particular , the gaussian distribution @xmath265 has cm @xmath266 , as evident from the decomposition of the average ( thermal ) state@xmath267 where@xmath268=\frac{\exp\left ( \frac{-\mathbf{a}^{t}\mathbf{a}}{2(\mu-1)}\right ) } { 2\pi(\mu-1)}.\ ] ] in the complex notation , the amplitude of the remote coherent state is given by @xmath269 where @xmath19 is the complex outcome of the heterodyne detection . for large @xmath270 , we have @xmath271 and @xmath272 .
consider now the other type of epr state , i.e. , the one having @xmath273 . from eqs .
( [ px_for])-([cmcond_for ] ) , we see that the sign change in @xmath223 only affects the first moments in eq .
( [ xcond_for ] ) .
this means that mode @xmath5 is projected on coherent states as before , the only difference being the relation between their amplitude and the outcome of the heterodyne detection , which is now@xmath274 thus , heterodyning mode @xmath6 of an arbitrary epr state with @xmath275 projects mode @xmath5 into an ensemble of coherent states with coherent amplitudes given by eq .
( [ cohprepa ] ) or eq .
( [ coheprepa2 ] ) . in terms of entropy
there is no difference , since the entropy of a gaussian state depends only on the second - order statistical moments .
here we show how we can create an ensemble of squeezed states by detecting one mode of an epr state . to achieve this , the gaussian state @xmath229 in the measurement operator @xmath78 of eq .
( [ meas_operator ] ) must be the squeezed vacuum , with cm @xmath276 where @xmath81 .
let us apply this local detection to mode @xmath6 of an epr state with blocks @xmath277 and @xmath278 .
using eq .
( [ cmcond_for ] ) we derive @xmath279{cc}r & 0\\ 0 & r^{-1}\end{array } \right ) , ~r:=\frac{1+u\mu}{u+\mu}.\ ] ] this is clearly the cm of a squeezed state .
in particular , it is easy to check that for @xmath84 ( corresponding to homodyne @xmath280-detection ) , we get @xmath281 , while for @xmath85 ( which is homodyne @xmath282-detection ) we get @xmath283 .
the amount of squeezing we can generate on mode @xmath5 is limited by the amount of correlations in the epr state ( quantified by the parameter @xmath270 ) .
finally , note that for @xmath83 , we have heterodyne detection and we get @xmath284 as expected ( coherent states ) . finally , using eqs .
( [ px_for ] ) and ( [ xcond_for ] ) we can derive both the gaussian probability of the outcomes and the modulated mean value of the states in terms of these outcomes .
however , we omit here this calculation , being irrelevant for the computation of the entropy .
for the sake of completeness , we prove here the formulas ( [ px_for])-([cmcond_for ] ) for the remote preparation of gaussian states .
we start by introducing the partial - trace rule , which is a central tool in our demonstration .
consider a two - mode state @xmath7 with characteristic function @xcite @xmath285 $ ] , and a state @xmath229 of mode @xmath6 with characteristic function
@xmath286 $ ] .
we consider the partial trace of their product , i.e. , the operator @xmath287 $ ] , with associated characteristic function @xmath288(\alpha):=\mathrm{tr}[\hat{o}\hat{d}(\alpha)]$ ] .
this function can be computed via the partial - trace rule @xcite@xmath289(\alpha)=\int_{\mathbb{c}}\frac{d^{2}\beta}{\pi}\chi ( \alpha,-\beta)\chi_{0}(\beta ) .
\label{chiob}\ ] ] this is a simple extension of the trace rule for single - mode states @xmath290 and @xmath229 , which reads@xmath291 in particular , when @xmath290 and @xmath229 are gaussian with statistical moments @xmath292 and @xmath293 , it is straightforward to check that@xmath294 } { \sqrt{\det(\mathbf{v}+\mathbf{v}_{0})}}. \label{tracerulegaussian}\ ] ] let us start with the proof of eq .
( [ px_for ] ) .
we write@xmath295 , \label{eqpa}\ ] ] where @xmath296 is the reduced state of mode @xmath6 .
since @xmath12 and @xmath297 are single - mode gaussian states with statistical moments @xmath298 and @xmath299 , we can directly apply eq .
( [ tracerulegaussian ] ) to obtain eq .
( [ px_for ] ) . to compute the statistical moments
@xmath245 and @xmath246 of the conditional gaussian state @xmath300 , \ ] ] we apply the partial - trace rule to @xmath301 $ ] . here
it is convenient to adopt the cartesian decomposition @xmath228 and @xmath302 , and use the following real variables @xmath303{c}\hat{q}\\ \hat{p}\end{array } \right ) , ~\mathbf{\tilde{x}}:=\left ( \begin{array } [ c]{c}\operatorname{im}\alpha\\ -\operatorname{re}\alpha \end{array } \right ) , ~\mathbf{\tilde{y}}:=\left ( \begin{array } [ c]{c}\operatorname{im}\beta\\ -\operatorname{re}\beta \end{array } \right ) .\ ] ] in this notation the displacement becomes @xmath304 and the characteristic function of @xmath242 reads@xmath305=\frac{1}{4\pi p(\mathbf{k})}\chi\lbrack\hat { o}_{\mathbf{k}}](\mathbf{\tilde{x } } ) , \label{chiatr}\ ] ] where @xmath306(\mathbf{\tilde{x}})$ ] is the characteristic function of @xmath307 . using the partial - trace rule we get@xmath308(\mathbf{\tilde{x}})=\int_{\mathbb{r}^{2}}\frac{d^{2}\mathbf{\tilde{y}}}{\pi}\chi(\mathbf{\tilde{x}},-\mathbf{\tilde { y}})\chi_{0}(\mathbf{\tilde{y } } ) , \label{ptr}\ ] ] where @xmath309=e^{-\frac{1}{2}\mathbf{\tilde{y}}^{t}\mathbf{v}_{0}\mathbf{\tilde{y}}-i\mathbf{k}^{t}\mathbf{\tilde{y}}}$ ] and @xmath310\\ & = e^{-\frac{1}{2}\left ( \mathbf{\tilde{x}}^{t}\mathbf{a\tilde{x}+\tilde{y}}^{t}\mathbf{b\tilde{y}+2\tilde{y}}^{t}\mathbf{c}^{t}\mathbf{\tilde{x}}\right ) -i\mathbf{\bar{x}}_{a}^{t}\mathbf{\tilde{x}-}i\mathbf{\bar{x}}_{b}^{t}\mathbf{\tilde{y}}}.\end{aligned}\ ] ] by solving the gaussian integral @xcite in eq .
( [ ptr ] ) and replacing in eq .
( [ chiatr ] ) , we obtain @xmath311 where the statistical moments , @xmath245 and @xmath246 , are those in eqs .
( [ xcond_for ] ) and ( [ cmcond_for ] ) .
in order to prove eq .
( [ chiob ] ) , we first express the states in terms of their characteristic functions @xmath312 then , we may write@xmath313 \\ & \times\chi(\alpha,\beta^{\prime})\chi_{0}(\beta)\hat{d}(-\alpha).\end{aligned}\ ] ] since@xmath314=\pi\delta ^{(2)}(-\beta^{\prime}-\beta)$ ] , we get@xmath315 \hat { d}(-\alpha),\ ] ] where the term inside the square brackets is @xmath288(\alpha)$ ] . | in this letter we exploit the recently - solved conjecture on the bosonic minimum output entropy to show the optimality of gaussian discord , so that the computation of quantum discord for bipartite gaussian states can be restricted to local gaussian measurements .
we prove such optimality for a large family of gaussian states , including all two - mode squeezed thermal states , which are the most typical gaussian states realized in experiments .
our family also includes other types of gaussian states and spans their entire set in a suitable limit where they become choi - matrices of gaussian channels . as a result
, we completely characterize the quantum correlations possessed by some of the most important bosonic states in quantum optics and quantum information .
quantum correlations represent a fundamental resource in quantum information and computation @xcite .
if we restrict the description of a quantum system to pure states , then quantum entanglement is synonymous with quantum correlations .
however , this is not exactly the case when general mixed states are considered : separable mixed states can still have residual correlations which can not be simulated by any classical probability distribution @xcite .
these residual quantum correlations are today quantified by quantum discord @xcite .
quantum discord is defined as the difference between the total correlations within a quantum state , as measured by the quantum mutual information , and its classical correlations , corresponding to the maximal randomness which can be shared by two parties by means of local measurements and one - way classical communication @xcite .
this definition not only provides a more precise characterization of quantum correlations but also has direct application in various protocols , including quantum state merging @xcite , remote state preparation @xcite , discrimination of unitaries @xcite , quantum channel discrimination @xcite , quantum metrology @xcite and quantum cryptography @xcite . for bosonic systems , like the optical modes of the electromagnetic field ,
it is therefore crucial to compute the quantum discord of gaussian states @xcite . despite these states being the most common in experimental quantum optics and the most studied in continuous - variable quantum information @xcite ,
no closed formula is yet known for their quantum discord .
what is computed is an upper - bound , known as gaussian discord @xcite , which is a simplified version based on gaussian detections only .
gaussian discord has been conjectured to be the actual discord for gaussian states , as also supported by recent numerical studies @xcite . in this letter
, we connect this conjecture on gaussian discord with the recently - solved conjecture on the bosonic minimum output entropy @xcite , according to which the von neumann entropy at the output of a single - mode gaussian channel is minimized by a pure gaussian state at the input @xcite .
in particular , this optimal input state is the vacuum , or any other coherent state , when we consider gaussian channels whose action is symmetric in the quadratures ( phase - insensitive ) , as for instance is the case of lossy or amplifier channels ( besides @xcite see also @xcite for an alternate proof ) . we show that the minimization of the bosonic output entropy implies the optimality of gaussian discord for a large family of gaussian states .
this family includes the important class of squeezed thermal states , which are all those states realized by applying two - mode squeezing to a pair of single - mode thermal states @xcite .
a key point in our analysis is showing that these states can always be decomposed into an einstein - podolsky - rosen ( epr ) state plus the local action of a phase - insensitive gaussian channel .
given such decomposition , we can easily show that heterodyne detection represents the optimal local measurement for computing their quantum discord .
more generally , by extending the previous decomposition to include other forms of local gaussian channels , we show that we can generate many other types of bipartite gaussian states , for which the optimal local measurement is gaussian , given by a ( quasi-)projection on single - mode pure squeezed states .
furthermore , our decomposition spans the entire set of gaussian states in a suitable ( and fastly - converging ) limit where they become choi - matrices of local gaussian channels . as a result of our study , we are now able to compute the actual unrestricted discord of a large portion of gaussian states , paving the way for a complete and precise characterization of the most fundamental quantum correlations possessed by bosonic systems . _ quantum discord and its gaussian formulation_. in classical information theory , the mutual information between two random variables , @xmath0 and @xmath1 , can be written as @xmath2 , where @xmath3 is the shannon entropy of variable @xmath0 , and @xmath4 is its conditional shannon entropy .
this notion has several inequivalent generalizations in quantum information theory @xcite , where the two variables are replaced by two quantum systems , @xmath5 and @xmath6 , in a joint quantum state @xmath7 . a first generalization is given by the quantum mutual information @xcite , defined as @xmath8 , where @xmath9 is the von neumann entropy of system @xmath5 , in the reduced state @xmath10 , and @xmath11 is the conditional von neumann entropy ( to be computed from the joint state @xmath7 and the other reduced state @xmath12 ) .
the quantum mutual information is a measure of the total correlations between the two quantum systems .
a second generalization is given by@xmath13 where the conditional term @xmath14 is the von neumann entropy of system @xmath5 minimized over all possible measurements on system @xmath6 , generally described as positive operator valued measures ( povms ) @xmath15 ( in particular , it is sufficient to consider rank-1 povms @xcite ) .
mathematically , this conditional term is written as @xmath16 where @xcite@xmath17 with @xmath18 the probability of outcome @xmath19 , and @xmath20 the conditional state of @xmath5 . the entropic quantity @xmath21 quantifies the classical correlations in the joint state @xmath7 , being the maximum amount of common randomness which can be extracted by local measurements and one - way classical communication @xcite .
quantum discord is then defined as the difference between the total and these classical correlations @xcite@xmath22 when the two systems @xmath5 and @xmath6 are bosonic modes , we can consider their gaussian discord @xmath23 , where the minimization of the conditional term @xmath14 in eq .
( [ mincond ] ) is restricted to gaussian povms .
thanks to this restriction , gaussian discord is easy to compute for two - mode gaussian states @xcite . furthermore , as we have already mentioned , gaussian discord is conjectured to be optimal for these states , in the sense that it would represent their unrestricted quantum discord , i.e. , @xmath24 .
this conjecture is supported by numerical studies @xcite and known to be true for a very limited set of gaussian states @xmath7 , namely those which can be purified into a three - mode gaussian state @xmath25 symmetric in the @xmath26 subsystem @xcite . _
normal forms , two - mode squeezed thermal states , and their decomposition_. since quantum discord and classical correlations are entropic quantities , they are invariant under local unitaries .
this means that we may apply displacements to yield a zero mean value , and local gaussian unitaries to reduce the covariance matrix ( cm ) into normal form @xcite .
thus , without loss of generality , quantum discord can be studied for zero - mean gaussian states @xmath7 with cm@xmath27{cc}% a\mathbf{i } & \mathrm{diag}(c , c^{\prime})\\ \mathrm{diag}(c , c^{\prime } ) & b\mathbf{i}% \end{array } \right ) : = \mathbf{v}(a , b , c , c^{\prime } ) , \label{normalform}%\ ] ] where @xmath28 and the parameters satisfy bona - fide conditions imposed by the uncertainty principle @xcite . for simplicity ,
we start from the most typical zero - mean gaussian states , i.e. , two - mode squeezed thermal states .
these states have cms of the form @xmath29 with bona - fide conditions @xmath30 and @xmath31 . as proven in the supplemental material and depicted in fig .
[ fig ] , these states can always be decomposed as @xmath32 , where @xmath33 is a phase - insensitive gaussian channel ( details below ) , @xmath34 is the identity channel , and @xmath35 is an epr state with cm@xmath36{cc}% b\mathbf{i } & \sqrt{b^{2}-1}\mathbf{c}\\ \sqrt{b^{2}-1}\mathbf{c } & b\mathbf{i}% \end{array } \right ) , \label{cmgene}%\ ] ] where@xmath37{cc}% \mathrm{sign}(c ) & 0\\ 0 & -\mathrm{sign}(c ) \end{array } \right ) .\ ] ] can always be decomposed into an epr state @xmath35 plus the local application of a phase - insensitive gaussian channel @xmath33 .
_ remote preparation _ ( red elements , see next section ) . a local measurement @xmath38 on mode @xmath6 generates a remote ensemble @xmath39 on the output mode @xmath5
. there will be another ensemble @xmath40 generated on the input mode @xmath41 before the channel .
this input ensemble will be made by gaussianly - modulated coherent states if @xmath38 is heterodyne detection.,scaledwidth=75.0% ] at the level of the second order moments , a phase - insensitive gaussian channel @xmath33 performs the transformation @xmath42 , with transmission matrix @xmath43 and noise matrix @xmath44 , where @xmath45 and @xmath46 .
in particular , this is a ` lossy channel ' for transmissivity @xmath47 $ ] and thermal noise @xmath48 , an ` additive - noise channel ' for @xmath49 and @xmath50 , or an ` amplifier channel ' for @xmath51 and @xmath52 .
( these are the most important canonical forms of a single - mode gaussian channel , see @xcite and supplemental material for this classification . ) as a result , the cm of a two - mode squeezed thermal state @xmath53 can be expressed in the equivalent form@xmath27{cc}% ( \tau b+\eta)\mathbf{i } & \sqrt{\tau(b^{2}-1)}\mathbf{c}\\ \sqrt{\tau(b^{2}-1)}\mathbf{c } & b\mathbf{i}% \end{array } \right ) , \label{cmab}%\ ] ] for some choice of @xmath45 and @xmath54
. _ remote state preparation and connection with the bosonic minimum output entropy_. as depicted in fig .
[ fig ] , the action of a local povm @xmath55 on mode @xmath6 generates a remote ensemble of states @xmath40 on the input mode @xmath41 , and a corresponding ensemble @xmath39 on the output mode @xmath5 . with probability @xmath56
, we have a conditional input state @xmath57 and its corresponding channel output @xmath58 .
assuming heterodyne detection @xmath59 , the input ensemble @xmath40 consists of coherent states @xmath60 whose complex amplitudes @xmath61 are gaussianly - distributed ( see supplemental material for more details on the remote preparation of gaussian states ) .
as a result , the output ensemble @xmath39 will be composed of gaussian states @xmath62 with gaussianly - modulated first moments and cm equal to @xmath63 .
the output entropy associated with the heterodyne detection is equal to the average entropy of the output ensemble @xmath39 , i.e.,@xmath64.\ ] ] since entropy is invariant under displacements , we may write @xcite @xmath65=s[\mathcal{e}% ( |0\rangle\langle0|)]$ ] and therefore@xmath66=h(\tau+\eta),\ ] ] where @xmath67 at this point we can exploit the solved conjecture on the bosonic minimum output entropy , which states that the vacuum ( or any other coherent state ) minimizes the output entropy of the phase - insensitive gaussian channel @xmath33 among all possible input states @xcite @xmath68=\inf_{\rho}s[\mathcal{e}(\rho ) ] . \label{minout}%\ ] ] as a result ,
we may write@xmath69\nonumber\\ & \leq\inf_{\mathcal{m}_{b}}s(a|\mathcal{m}_{b})=s_{\min}(a|b),\end{aligned}\ ] ] where the inequality comes from the fact that any @xmath38 , with input ensemble @xmath70 and output ensemble @xmath71 , must satisfy @xmath72\geq\inf_{\rho}% s[\mathcal{e}(\rho)].\end{aligned}\ ] ] thus , there is a gaussian povm ( heterodyne detection ) which is optimal for the minimization of the output entropy @xmath73 .
this is equivalent to saying that the gaussian discord of the gaussian state @xmath7 is optimal , i.e. , equal to its actual discord .
its calculation is therefore easy , since @xmath74 , leading to @xmath75 where @xmath76 is the symplectic spectrum of @xmath77 , which can be easily computed @xcite from eq .
( [ cmab ] ) .
_ extending the family of gaussian states_. here we extend the previous derivation to include other gaussian states .
we first generalize the local gaussian povm @xmath38 , whose element @xmath78 becomes a quasi - projector on the squeezed state @xmath79
@xcite with variable amplitude @xmath19 but fixed cm @xmath80 , where @xmath81 .
this measurement @xmath82 corresponds to a heterodyne detection for @xmath83 , and becomes a homodyne detection for @xmath84 or @xmath85 . by applying @xmath86 to an epr state @xmath35 with variance @xmath87 as in eq .
( [ cmgene ] ) , we generate an ensemble @xmath40 of amplitude - modulated squeezed states with cm @xmath88 , where @xmath89 .
the value of this squeezing ranges from @xmath90 to @xmath91 , extremes which are achieved by the two homodyne detectors ( see supplemental material for more details ) . which can be decomposed into an epr state @xmath35 by locally applying a gaussian channel @xmath33 plus input - output squeezing operators .
here the local detection @xmath38 is a quasi - projection onto squeezed states , so that the input ensemble @xmath40 is composed of amplitude - modulated squeezed states.,scaledwidth=56.0% ] we may now rectify the ensemble @xmath40 by applying the anti - squeezing operator @xcite @xmath92 which transforms its states into coherent states ( see fig .
[ decopic ] ) . in this way
we are sure that optimal states are fed into the gaussian channel @xmath33 , whose output entropy is therefore minimized . furthermore , this output entropy does not change if we apply another squeezing operator @xmath93 , with a suitable @xmath94 putting the gaussian state @xmath7 into normal form .
thus the optimality of gaussian discord is proven for any gaussian state which is decomposable into an epr state as @xmath95 with optimal detection given by the gaussian @xmath86 .
the second generalization consists of extending the gaussian channel @xmath33 to include negative transmissivities @xmath96 , where the channel describes the conjugate of an amplifier , which is another gaussian channel ( phase - sensitive ) whose output entropy is minimized by coherent states at the input @xcite .
we can then consider an extended gaussian channel @xmath33 with arbitrary @xmath97 and @xmath98 , which is described by the matrices @xmath99 $ ] and @xmath100 .
thus , the optimality of gaussian discord is proven for all gaussian states @xmath7 decomposable as in eq .
( [ gendec ] ) with @xmath33 being such an extended channel .
as we show in the supplemental material , these states have cms in the normal - form @xmath101 , where@xmath102\sqrt{\left\vert \tau\right\vert ( b^{2}-1)\theta(r)/\theta(r^{-1 } ) } , \label{cp}%\end{aligned}\ ] ] with @xmath97 , @xmath46 and @xmath103 $ ] . here
the sign ambiguity comes from the type of epr state considered in the decomposition , whose cm in eq .
( [ cmgene ] ) is generally defined with @xmath104 .
also note that negative transmissivities allow us to include states with @xmath105 . for arbitrary fixed values of @xmath106 and @xmath107
, we generate all accessible values of the correlation parameters @xmath108 and @xmath109 by exploiting the remaining degrees of freedom in the parameters @xmath110 , @xmath111 and @xmath112 . as shown in the numerical investigation of fig .
[ pictot ] , our family represents a wide portion of all gaussian states .
it includes all states whose cms have the form @xmath101 with @xmath113 , corresponding to the bisectors of the correlation plane @xmath114 .
furthermore , for increasing @xmath87 , our family tends to invade the entire gaussian set very quickly .
this set is completely filled for @xmath115 , where the epr state @xmath35 becomes maximally entangled and the gaussian state @xmath7 becomes the choi matrix of the gaussian channel @xmath116 . by increasing the entanglement in the epr state , we increase the amount of squeezing @xmath112 that we can generate in the input ensemble @xmath40 .
the effect of this squeezing is to include output states where @xmath117 and @xmath118 are very different , as also evident from the @xmath112-dependence in eqs .
( [ c ] ) and ( [ cp ] ) .
in particular , in the limit of @xmath119 , homodyne detectors generate infinite squeezing in @xmath40 and we approach all the exotic states on the axes of the plane , for which one of the correlation parameters is zero ( @xmath120 ) . and
@xmath87 , the correlation parameters @xmath108 and @xmath109 can only take a restricted range of physical values corresponding to the delimited regions in the panels . within these regions
the darker points are members of our gaussian family .
plots are created by randomly testing @xmath121 points.,scaledwidth=49.0% ] _ conclusion and discussion_. in this letter , we have shown that the validity of the bosonic minimum output entropy conjecture implies the optimality of gaussian discord for a large family of gaussian states , decomposable into epr states subject to local gaussian channels .
in particular , this family includes the class of two - mode squeezed thermal states @xcite . from this point of view
, our work completely characterizes the quantum correlations possessed by the most typical states in continuous variable quantum information and experimental quantum optics .
the exact size of these quantum correlations , i.e. , their unconstrained quantum discord , can now be computed efficiently ( note that computing discord is np - complete in the general case @xcite ) .
we have shown that we can rapidly fill the entire set of gaussian states by increasing their parameter @xmath87 , i.e. , the thermal variance in the mode under detection .
the reason is because this parameter corresponds to the variance of the epr state involved in the decomposition . by increasing this parameter
, we increase the epr entanglement and therefore the amount of squeezing that we can remotely generate at the input of the gaussian channel .
we could further improve our results if an energy - constrained version of the entropy conjecture were proven for the ` pathological ' canonical forms @xcite .
in fact , there exist highly phase - sensitive gaussian channels which completely destroy the correlations in only one of the quadratures , therefore being particularly suitable to decompose exotic gaussian states with @xmath120 .
unfortunately , no finite - energy state is known to be optimal for the minimization of the output entropy of these channels .
if such a state were gaussian , then it could be prepared with a limited amount of squeezing and we would span the entire set of gaussian states more straightforwardly . finally , we remark that the complete characterization of the quantum correlations possessed by gaussian states is important not only in quantum information and quantum optics ( e.g. , for problems of quantum metrology ) , but also in other fields .
these include condensed matter physics ( e.g. , bose - einstein condensates ) , solid - state physics , relativistic quantum field theory ( where gaussian states arise from bogoliubov transformations , e.g. , in the unruh effect or the hawking radiation ) , statistical mechanics and foundations of quantum mechanics . _
acknowledgments_. this work was funded by a leverhulme trust research fellowship and epsrc ( grants ep / j00796x/1 and ep / l011298/1 ) .
s.p . would like to thank m.g.a .
paris for enlightening discussions .
99 m.a .
nielsen and i.l .
chuang , _ quantum computation and quantum information _
( cambridge university press , cambridge , 2000 ) . m.m .
wilde , _ quantum information theory _
( cambridge university press , cambridge 2013 ) .
h. ollivier and w.h .
zurek , phys .
rev .
lett . * 88 * , 017901 ( 2001 ) .
l. henderson and v. vedral , j. phys .
a * 34 * , 6899 ( 2001 ) .
k. modi , a. brodutch , h. cable , t. paterek , and v. vedral , rev .
mod .
phys . * 84 * , 1655 - 1707 ( 2012 ) .
i. devetak and a. winter , ieee trans .
inform . theory .
* 50 * , 3183 , ( 2004 ) .
d. cavalcanti , l. aolita , s. boixo , k. modi , m. piani , a. winter , phys .
rev .
a * 83 * , 032324 ( 2011 ) ; v. madhok and a. datta , phys .
rev .
a * 83 * , 032323 ( 2011 ) .
b. dakic , y.o .
lipp , x. ma , m. ringbauer , s. kropatschek , s. barz , t. paterek , v. vedral , a. zeilinger , c. brukner , and p. walther , nature phys .
* 8 * , 666 - 670 ( 2012 ) .
m. gu , h.m .
chrzanowski , s.m .
assad , t. symul , k. modi , t.c .
ralph , v. vedral , and p. koy lam , nature phys . * 8 * , 671 - 675 ( 2012 ) .
c. weedbrook , s. pirandola , j. thompson , v. vedral , and m. gu , preprint arxiv:1312.3332 . d. girolami , t. tufarelli , and g. adesso , phys .
rev .
lett . * 110 * , 240402 ( 2013 ) ; d. girolami _ et al_. , phys .
rev .
lett . * 112 * , 210401 ( 2014 ) .
s. pirandola , sci . rep .
* 4 * , 6956 ( 2014 ) . c. weedbrook , s. pirandola , r. garcia - patron , n.j .
cerf , t.c .
ralph , j.h .
shapiro , and s. lloyd , rev .
mod .
phys . * 84 * , 621 ( 2012 ) .
s.l . braunstein and p. van loock , rev .
mod .
phys . * 77 * , 513 ( 2005 ) .
p. giorda and m.g.a .
paris , phys .
rev .
lett . * 105 * , 020503 ( 2010 ) .
g. adesso and a. datta , phys .
rev .
lett . * 105 * , 030501 ( 2010 ) .
p. giorda , m. allegra , and m.g.a .
paris , phys .
rev .
a * 86 * , 052328 ( 2012 ) .
s. olivares and m.g.a .
paris , int .
j. mod .
phys .
b * 27 * , 1245024 ( 2012 ) .
v. giovannetti , r. garcia - patron , n.j .
cerf , and a.s .
holevo , nature photonics * 8 * , 796 ( 2014 ) .
a. mari , v. giovannetti , and a.s .
holevo , nature comm .
* 5 * , 3826 ( 2014 ) .
note that the sum can be over a continuous parameter , in which case it must be interpreted as an integral . by applying the koashi - winter relation @xcite to the pure state @xmath25 , we may write @xmath122 , where @xmath123 is the entanglement of formation of the two - mode gaussian state @xmath124 .
if @xmath125 , then @xmath123 can easily be computed as gaussian entanglement of formation @xcite and this calculation confirms that @xmath14 can be achieved by gaussian povms .
m. koashi , and a. winter , phys .
rev .
a * 69 * , 022309 ( 2004 ) .
m.m .
wolf , g. giedke , o. krueger , r.f .
werner , and j.i .
cirac , phys .
rev .
a * 69 * , 052320 ( 2004 ) .
s. pirandola , a. serafini , and s. lloyd , phys .
rev .
a * 79 * , 052327 ( 2009 ) .
s. pirandola , new j. phys .
* 15 * , 113046 ( 2013 ) here we use @xmath126 , so that the shot - noise variance of the vacuum is equal to one .
any single - mode gaussian channel can be dilated into a gaussian unitary involving the input mode @xmath127 plus two environmental modes @xmath128 and @xmath129 .
globally the transformation of the quadratures @xmath130 is an affine map @xmath131 , where @xmath132 is a symplectic matrix and @xmath133 is a displacement vector associated with the channel . now
any displacement of the input mode @xmath134 , with @xmath135 will be mapped into a corresponding ( different ) displacement of the output mode @xmath136 . since entropy is invariant under unitaries , it is not affected by these output displacements and , therefore , by the input ones .
y. huang , new j. phys .
* 16 * , 033027 ( 2014 ) .
a.s .
holevo , probl .
inf .
transm .
* 43 * , 1 ( 2007 ) .
s. pirandola , s.l .
braunstein , and s. lloyd , phys .
rev .
lett .
* 101 * , 200504 ( 2008 ) . supplemental material * contents of the document*. in this supplemental material , we start by giving a brief review on single - mode gaussian channels and their canonical forms , also discussing the minimization of their output entropy ( sec . [ canforms_app ] ) . in sec .
[ deco1_app ] we prove the decomposition of two - mode squeezed thermal states . in sec .
[ deco2_app ] we prove the parametrization of our general family of gaussian states and we describe the random sampling of this family in more detail . in sec .
[ remote_app ] , we provide the mathematical tools for computing the remote preparation of gaussian states by applying a local gaussian measurement to an arbitrary two - mode gaussian state .
formulas will be then specified to the case of the epr state , subject to the detections considered in the main article . because some of these tools are only partially known in the literature , we provide a comprehensive proof in the final sec .
[ appmiscellanea ] . |
a _ fibred link _ is a link @xmath5 such that @xmath6 fibers over the circle , and where each fibre is the interior of a seifert surface @xmath7 for @xmath8 in @xmath9 . cutting @xmath7 along a properly embedded interval @xmath10 ( an _ arc _ for short ) results in another seifert surface @xmath11 for another link @xmath12 .
if @xmath13 is again a fibred link with fibre @xmath11 , we say that @xmath10 _ preserves fibredness_. for example , @xmath10 could be the spanning arc of a plumbed hopf band , and cutting along @xmath10 amounts to deplumbing that hopf band . in @xcite ,
buck et al .
give a simple criterion for when an arc preserves fibredness in terms of the monodromy @xmath14 .
as a corollary , they prove that each of the torus links of type @xmath15 admits only a finite number of such arcs up to isotopy .
it turns out that among torus links , this is an exception : [ thm:1 ] let @xmath16 or @xmath17 .
then the fibre surface @xmath7 of the torus link @xmath18 contains infinitely many homologically distinct cutting arcs preserving fibredness .
the remaining torus links @xmath15 , @xmath19 , @xmath20 and @xmath21 happen to be exactly those torus links that can also be obtained as plumbings of positive hopf bands according to a finite tree , where vertices correspond to positive hopf bands and edges indicate plumbing .
[ thm:2 ] let @xmath7 be the fibre surface obtained by plumbing positive hopf bands according to a finite tree @xmath22 .
there are , up to isotopy , only finitely many cutting arcs in @xmath7 preserving fibredness , if and only if @xmath22 is one of the coxeter - dynkin trees @xmath0 , @xmath1 , @xmath2 , @xmath3 or @xmath4 . to prove the `` only if '' part of theorem [ thm:2 ] , we consider orbits of a fixed arc under the monodromy to produce families of arcs that preserve fibredness .
the basic idea is that such an orbit is infinite if the monodromy has infinite order .
for example , we show that in fact every ( prime ) positive braid link with pseudo - anosov monodromy admits infinitely many non - isotopic arcs preserving fibredness .
this suggests the following question : is it true that among all ( non - split prime ) positive braid links , the ade plane curve singularities are exactly those that admit just a finite number of fibredness preserving arcs up to isotopy ?
we use the shorthand _
@xmath23 links _ to refer to the links of the positive tree - like hopf plumbings according to the trees @xmath0 ,
@xmath1 , @xmath2 , @xmath3 or @xmath4 .
the subsequent section combines a criterion on arcs to preserve fibredness from @xcite with the property of monodromies of positive hopf plumbed surfaces to be right - veering .
this allows for the following simple test for an arc to preserve fibredness , in our situation : an arc preserves fibredness if and only if it does not intersect its image under the monodromy ( up to free isotopy ) .
section [ sec : monodromies ] contains descriptions of the fibre surfaces and the monodromies of the links we consider ( torus links and the @xmath23 links ) . alongside ,
we give a constructive proof of theorem [ thm:1 ] . in section [ sec : exceptional ] , we explain the idea of proof for the finiteness result that provides the `` if '' part of theorem [ thm:2 ] , and list the fibred links obtained by cutting the fibre surfaces of the @xmath23 links along an arc in table [ table ] . section [ sec : infinite ] accounts for the cases where the monodromy has infinite order .
this concerns in particular the positive tree - like hopf plumbings that correspond to trees different from the @xmath23 trees and settles the `` only if '' part of theorem [ thm:2 ] . at the beginning of section [ sec : propproofs ] , we set up the notation and methods needed for the proof of the finiteness part of theorem [ thm:2 ] , which we split into proposition [ prop:3345 ] ( concerning torus links ) and proposition [ prop : dne7 ] ( concerning tree - like hopf plumbings ) .
the rest of that section is devoted to the proofs of these propositions .
in the sequel we would like to make statements on the relative position of two arcs @xmath24 in a surface @xmath7 with boundary ( that is , @xmath24 are embedded intervals with endpoints on the boundary of @xmath7 that are nowhere tangent to @xmath25 ) .
the following definition will simplify matters .
let @xmath7 be an oriented surface with boundary and let @xmath26 be two arcs .
a property @xmath27 is said to hold _
after minimising isotopies on @xmath10 and @xmath28 _ , if @xmath29 holds , where @xmath30 and @xmath31 are obtained from @xmath10 , @xmath28 by two isotopies ( fixed at the endpoints ) that minimise the geometric number of intersections between the two arcs .
the remainder of this section will recall the fact that every positive braid link ( that is , the closure of a braid word consisting only of the positive generators of the braid group , without their inverses ) is fibred and has so - called _ right - veering _ monodromy ( see below for a definition ) .
the torus links @xmath18 provide examples , since they can be viewed as the closures of the positive braids @xmath32 , where the @xmath33 denote the ( positive ) standard generators of the braid group .
let @xmath7 be an oriented surface with boundary and @xmath34 a diffeomorphism that restricts to the identity on @xmath25 .
then , @xmath35 is called _ right - veering _ if for every arc @xmath36\to s$ ] , the vectors @xmath37 form an oriented basis after minimising isotopies on @xmath10 and @xmath38 .
this means basically that arcs starting at a boundary point of @xmath7 get mapped `` to the right '' by @xmath35 .
it is known that every positive braid can be obtained as an iterated plumbing of positive hopf bands ( see @xcite ) . since
a hopf band is a fibre and plumbing preserves fibredness , every positive braid link is fibred .
moreover the monodromy is a product of positive dehn twists , since the monodromy of a ( positive ) hopf band is a ( positive ) dehn twist and the monodromy of a plumbing is the composition of the monodromies of the plumbed surfaces ( see @xcite ) .
a product of positive dehn twists is right - veering @xcite .
so we conclude that every positive braid link is fibred with right - veering monodromy . together with a theorem by buck et al .
, this property implies the following simple geometric criterion for when an arc preserves fibredness . ]
[ thm : rv - cutting - arcs ] let @xmath8 be a fibred link with fibre surface @xmath7 and right - veering monodromy @xmath39 .
then , a cutting arc @xmath10 preserves fibredness if and only if @xmath40 after minimising isotopies on @xmath10 and @xmath41 .
this is a special case of theorem 1 in @xcite , saying that the arc @xmath10 preserves fibredness if and only if @xmath10 is _ clean and alternating _ or _ once unclean and non - alternating _ ( see figure [ fig : alternating ] ) , without the assumption on @xmath35 to be right - veering .
but for a right - veering map , every arc is alternating , by definition . finally , @xmath10 is clean if and only if @xmath40 after minimising isotopies on @xmath10 and @xmath41 .
[ rem : clean ] an arc @xmath10 is clean if and only if @xmath42 is clean , for all @xmath43 .
this is clear since @xmath40 after minimising isotopies if and only if @xmath44 after minimising isotopies .
similarly , if @xmath45 is a homeomorphism such that @xmath46 , then @xmath10 is a clean arc if and only if @xmath47 is .
indeed , @xmath48 .
the links that correspond to the trees @xmath0 , @xmath2 and @xmath4 are torus links , namely @xmath49 , @xmath50 and @xmath51 . together with @xmath52 , which is @xmath19 , these form the intersection between torus links and positive tree - like hopf plumbings . for our purpose
it therefore suffices to study torus links , @xmath3 and the @xmath1 family .
the monodromies @xmath39 of the links in question are particular examples of _ twists _ , a notion invented by acampo and further developed by graf in his thesis @xcite .
this means that there exists a @xmath35-invariant spine @xmath53 , called _
graph_. cutting @xmath7 along the graph results in finitely many annuli , on which @xmath35 descends to certain twist maps .
more precisely , each of these annuli has one component of @xmath25 as one boundary circle and a cycle consisting of edges of @xmath54 as the other .
@xmath35 fixes @xmath25 pointwise and rotates the edge - cycles by some number @xmath55 of edges .
the number @xmath56 is called the _ twist length _ of the corresponding boundary annulus . after an isotopy ( fixing the boundary of @xmath7 ) , we may therefore assume that @xmath35 is periodic except on some annular neighbourhoods of @xmath25 .
it is thus easy to understand the effect of @xmath35 on an arc @xmath10 , up to isotopy , given the combinatorics of the action of @xmath35 on @xmath54 and the amount of twisting on each annulus .
note that twists define periodic mapping classes in the sense that some power is freely isotopic to the identity .
however , this isotopy can not be taken to be fixed on the boundary of @xmath7 . in a way dual to the graph
, we will find in each case a finite set of disjoint arcs that are permuted by @xmath35 and which decompose @xmath7 into finitely many polygons , one for each vertex of @xmath54 .
the combinatorics of how these polygons are permuted will be used to prove theorems [ thm:1 ] and [ thm:2 ] .
the fibre surface @xmath7 of the torus link @xmath18 can be constructed as thickening of a complete bipartite graph on @xmath57 and @xmath58 vertices in the following way , as in figure [ fig : knm ] . and
@xmath59 vertices and blackboard framed thickening.[fig : knm ] ] arrange @xmath57 collinear points @xmath60 ( in this order ) in a plane and , similarly , another @xmath58 points @xmath61 along a line parallel to the @xmath62 .
connect @xmath62 and @xmath63 by a straight segment @xmath64 , for every @xmath65 , @xmath66 .
avoid intersections between the segments by letting @xmath64 pass slightly under @xmath67 if @xmath68 and @xmath69 ( in a slight thickening of the plane containing the points @xmath62 and @xmath63 ) .
use the blackboard - framing to thicken @xmath62 , @xmath63 , @xmath64 to disks @xmath70 , @xmath71 and bands @xmath72 . choose the thickness of the bands
@xmath72 so that they do not intersect outside the disks @xmath73 . it can be seen that @xmath74 is isotopic to the minimal seifert surface of @xmath18 in @xmath9 ( compare @xcite ) .
in addition , the monodromy @xmath39 is a twist along the above graph . in each of the @xmath75 complementary annuli , @xmath35 fixes @xmath25 pointwise and rotates the edge - cycles two edges to the right with respect to the orientation of @xmath7 . using this description , it is possible to see that @xmath35 acts on the graph as follows : @xmath76 , @xmath77 , @xmath78 , where the indices @xmath79 are to be taken modulo @xmath80 respectively .
a subarc of @xmath10 that travels near @xmath64 will be mapped to a subarc of @xmath41 that travels near @xmath81 .
the edges @xmath64 induce a decomposition of @xmath82 into circular arcs lying between points of the form @xmath83 ( and the same for @xmath71 ) . if @xmath84 , it is hence meaningful to speak of points on @xmath82 _ between _ @xmath64 and @xmath85 .
let @xmath16 or @xmath17 .
then the fibre surface @xmath7 of the torus link @xmath18 contains infinitely many homologically distinct cutting arcs preserving fibredness .
( solid line ) and its image under the monodromy ( dotted line ) .
note that these two arcs do not intersect , except at their endpoints.[fig : infiniteintervalnm ] ] for @xmath16 consider the following arcs in @xmath7 , using the notation from above ( compare figure [ fig : infiniteintervalnm ] ) : * let @xmath86 be a straight segment starting at a point of @xmath87 between @xmath88 and @xmath89 and ending at the vertex @xmath90 .
* let @xmath91 start at @xmath90 , follow the edges @xmath92 and @xmath93 , thus ending at @xmath94 .
* @xmath95 starts at @xmath94 , runs along @xmath96 , @xmath88 , @xmath92 , @xmath93 and ends again at @xmath94 .
* @xmath97 is a straight segment from @xmath94 to a point of @xmath98 between @xmath96 and @xmath99 . from @xmath100
we can build an infinite family @xmath101 of arcs in @xmath7 , taking @xmath102 . here
, @xmath103 denotes concatenation of paths . replacing the @xmath104 consecutive copies of @xmath95 by @xmath104 parallel copies
, the @xmath105 can be thought of as embedded arcs .
it is now easy to check that @xmath105 and its image @xmath106 under the monodromy @xmath35 have only their endpoints in common . using theorem [ thm : rv - cutting - arcs ]
it follows that each @xmath105 preserves fibredness
. finally , the @xmath105 are homologically pairwise distinct .
this can be seen in the following way : let @xmath107\in h_1(s,\z)$ ] be the cycle represented by a simple closed curve @xmath108 whose image is @xmath109 . after an isotopy , @xmath108 will intersect @xmath105 transversely in @xmath110 points .
now , the linear form on @xmath111 that sends @xmath10 to @xmath112 , the number of intersections with @xmath108 ( counted with signs ) , defines an element @xmath113 of @xmath114 such that @xmath115 , hence the claim . if @xmath17 , take the following arcs ( compare figure [ fig : infiniteinterval3 m ] ) : ( solid line ) and its image under the monodromy ( dotted line ) for a @xmath116 torus link , @xmath117 .
again , the two arcs do not intersect.[fig : infiniteinterval3 m ] ] * @xmath86 is a straight segment from a point of @xmath118 between @xmath119 and @xmath120 to @xmath121 .
* @xmath91 starts at @xmath121 , follows the edges @xmath122 and @xmath123 , thus ending at @xmath124 .
* @xmath95 starts at @xmath124 , follows @xmath125 , @xmath126 , @xmath127 , @xmath119 , @xmath122 and @xmath123 , ending at @xmath124 . *
@xmath97 is a straight segment from @xmath124 to a point of @xmath128 between @xmath129 and @xmath125 .
as above , we get a family @xmath101 of homologically distinct arcs preserving fibredness , where @xmath102 , using the curve with image @xmath130 to distinguish the @xmath105 . in order to obtain a similar model for the fibre surface @xmath7 of @xmath3 or @xmath1 , start with two disjoint planar disks @xmath131 in @xmath132 and
connect them by half twisted bands @xmath133 , where @xmath134 in the case of @xmath3 .
the embedded surface @xmath135 is then a fibre surface for the @xmath15 torus link .
let @xmath136 be a point between @xmath137 and @xmath138 in the case of @xmath1 , respectively between @xmath138 and @xmath139 in the case of @xmath3 .
let @xmath140 be an arc in @xmath141 from a point of @xmath142 between @xmath143 and @xmath144 to @xmath145 . finally , define @xmath7 to be the surface obtained from @xmath11 by plumbing a positive hopf band along @xmath140 below the surface @xmath11 .
denote the core curve of that plumbed hopf band by @xmath146 ( so @xmath147 ) .
each pair of consecutive bands @xmath148 , @xmath149 , gives rise to a closed curve @xmath150 that runs from @xmath141 to @xmath151 through @xmath152 and back to @xmath141 through @xmath153 .
fibre surface with homology basis coming from the plumbing tree .
@xmath35 is the product of the right handed dehn twists on the cuves @xmath154 .
[ fig : e7surface ] ] .
hexagon @xmath155 is shaded grey .
the monodromy permutes the intervals @xmath156 ( marked @xmath157 ) cyclically .
[ fig : e7intervals ] ] the incidence graph for the system of curves @xmath158 in @xmath7 is exactly the respective coxeter - dynkin tree @xmath3 or @xmath1 ( compare figure [ fig : e7surface ] ) .
the @xmath154 are core curves of positive hopf bands and @xmath7 is a tree - like positive hopf plumbing according to the respective tree . in particular ,
the monodromy @xmath35 of @xmath7 is the product of the right handed dehn twists about the curves @xmath159 , in this order . just as in the case of torus links ,
we will find a finite number of disjoint arcs in @xmath7 that are permuted ( up to free isotopy ) by @xmath35 and such that these arcs cut @xmath7 into polygons .
for @xmath3 , let @xmath160 be the spanning arc of @xmath161 , and let @xmath162 , @xmath163 , up to free isotopy ( compare figure [ fig : e7intervals ] ) . up to free isotopy , @xmath164 .
this can be seen by applying the dehn twists about the @xmath165 to the @xmath156 , as described above . another more visual way to see
this is via _ dragging arcs_. imagine the arcs @xmath156 to be elastic bands whose ends are attached to the surface boundary and whose interiors are pushed slightly off the surface into the positive normal direction . applying the monodromy @xmath35
amounts to dragging the arc through the complement of @xmath7 to the negative side of the surface , while its endpoints stay fixed on @xmath25 . since we are only interested in the position of @xmath166 up to free isotopy , the endpoints of the dragging arc may move freely along @xmath25 during that process .
let @xmath167 be the three disk components of @xmath168 .
the boundary of @xmath169 alternates between parts of @xmath25 and the @xmath156 .
we choose the order as in figure [ fig : e7hexagons ] , where the components of @xmath170 are shrunk to points . to @xmath171 , indices taken modulo @xmath59 , such that edge @xmath156 is sent to edge @xmath172 , modulo @xmath173 .
[ fig : e7hexagons ] ] examination of the action of @xmath35 on the @xmath156 reveals that the @xmath70 are cyclically permuted by @xmath35 , in the order @xmath174 . in figure
[ fig : e7hexagons ] , the @xmath70 are drawn in such a way that @xmath175 by translation to the right , and @xmath155 is mapped to @xmath176 by a translation , followed by a clockwise rotation through @xmath177 . to obtain the graph @xmath54 , put a vertex in the middle of each hexagon @xmath169 and connect them by edges through the center of every @xmath156 , connecting the vertices of the adjacent hexagons .
the twist lengths on the two boundary annuli are @xmath178 and @xmath179 , respectively . for the case of @xmath1 , @xmath57 odd , take @xmath160 to be the spanning arc of @xmath143 and let @xmath180 , @xmath181 . as before , we have @xmath182 , and the @xmath156 decompose @xmath7 into @xmath183 disks @xmath184 , as in figure [ fig : dnintervalsodd ] . in figure [ fig : dnsquares ] ( top ) , @xmath35 maps @xmath185 by right translations and sends @xmath186 back to @xmath176 by a rotation of @xmath187 . if @xmath57 is even , we use two orbits of intervals instead of one : define @xmath188 and @xmath189 by letting @xmath190 be the spanning arcs of @xmath143 , @xmath144 respectively and @xmath180 , @xmath191 .
again , the union of the @xmath156 and the @xmath192 decomposes @xmath7 into disks @xmath193 ( see figure [ fig : dnintervalseven ] ) . in figure [ fig : dnsquares ] ( bottom ) , the monodromy maps @xmath194 by translations .
the graphs for @xmath1 have one vertex at the center of each square and edges pass through the @xmath156 and @xmath192 .
twist lengths on the boundary annuli are @xmath178 , @xmath195 for odd @xmath57 , and @xmath178 , @xmath179 , @xmath196 for even @xmath57 . on the fibre surface of @xmath1 for even @xmath57.[fig : dnintervalseven ] ] on the fibre surface of @xmath1 for even @xmath57.[fig : dnintervalseven ] ] , for odd @xmath57 ( top ) and for even @xmath57 ( bottom).[fig : dnsquares],title="fig : " ] , for odd @xmath57 ( top ) and for even @xmath57 ( bottom).[fig : dnsquares],title="fig : " ]
in ( * ? ? ? * corollary 2 ) , buck et al . show that @xmath15 admits only finitely many arcs preserving fibredness ( up to isotopy ) .
more precisely , they show that every clean arc is isotopic ( free on the boundary ) to an arc that is contained in one of the disks @xmath197 from the above description of the monodromy of torus links . apart from this infinite family of torus links
, there are only three more torus links with just a finite number of arcs that preserve fibredness : the torus links @xmath19 , @xmath20 and @xmath21 admit , up to isotopy ( free on the boundary ) , only a finite number of cutting arcs that preserve fibredness . for positive tree - like
hopf plumbed surfaces we similarly obtain : the positive tree - like hopf plumbings associated to any of the coxeter - dynkin trees @xmath0 , @xmath1 , @xmath2 , @xmath3 or @xmath4 admit , up to isotopy ( free on the boundary ) , only a finite number of cutting arcs that preserve fibredness .
the proofs of propositions [ prop:3345 ] and [ prop : dne7 ] are rather technical and will be given in section [ sec : propproofs ] . nevertheless , the idea is very simple : let @xmath7 be the fibre surface of any torus link @xmath18 , given as thickening of a complete bipartite graph on @xmath198 vertices , or of @xmath1 or @xmath3 , as described in section [ sec : monodromies ] .
an arc @xmath199 is determined up to isotopy by its endpoints and by the sequence of bands @xmath200 it passes through .
now start listing all possible such sequences that yield clean arcs , for increasing length of the sequence . in order to prove finiteness of this list , we use three lemmas ,
also given in section [ sec : propproofs ] .
the intuitive meaning of lemma [ lem : conflict ] and lemma [ lem : generalisedconflict ] can be phrased as follows : if @xmath10 and @xmath41 intersect and this intersection seemingly can not be removed by an isotopy , then @xmath10 is indeed unclean .
lemma [ lem : consecutive ] asserts that a clean arc can not stay in the complement of the graph for a distance of more than @xmath55 consecutive bands , where @xmath55 is the twist length on the corresponding boundary annulus ( for example , @xmath201 for all torus links ) .
this is made precise in section [ sec : propproofs ] , using a notion of _ arcs in normal position _
definition [ def : normalpos ] ) . along with this case - by - case analysis
, one can find all possible fibred links obtained from @xmath202 , @xmath203 , @xmath1 , @xmath50 , @xmath3 and @xmath51 by cutting along an arc .
consult table [ table ] for a complete list .
.fibred links obtained from the exceptional torus links by cutting along an arc .
[ table ] [ cols="<,<",options="header " , ] @xmath204 denotes the connected sum of @xmath205 and @xmath206 , @xmath1 denotes the closure of the braid @xmath207 , @xmath208 , and @xmath209 denotes the closure of the braid @xmath210 , @xmath211 .
+ @xmath212 chain of four successive unknots .
+ @xmath213 both hopf link components are summed to one trefoil knot each .
+ @xmath214 both possible sums appear ( trefoil summed with the unknot component of @xmath215 as well as trefoil summed with the trefoil component of @xmath215 ) .
+ @xmath216 one component of the hopf link in the middle is summed to @xmath217 and the other is summed to the trefoil .
[ thm:3 ] let @xmath7 be a surface obtained by iterated plumbing of positive hopf bands and suppose the monodromy @xmath39 is pseudo - anosov . then , @xmath7 contains infinitely many non - isotopic cutting arcs preserving fibredness .
the monodromy @xmath35 is a composition of right dehn twists along the core curves of the hopf bands used for the construction of @xmath7 as a hopf plumbing .
let @xmath10 be an arc dual to the core curve of the last plumbed hopf band and such that @xmath10 does not enter any of the previously plumbed hopf bands .
then , in the product of dehn twists representing @xmath35 , only the last factor affects @xmath10 .
it follows that @xmath10 is clean ( and therefore @xmath218 is also clean by remark [ rem : clean ] ) . since @xmath35 is pseudo - anosov and @xmath10 is essential , the length of @xmath218 ( with respect to an auxiliary riemannian metric ) grows exponentially as @xmath57 tends to infinity ( compare @xcite , section 14.5 ) . in particular
, the arcs @xmath218 are pairwise non - isotopic and clean . in general
, it is not enough to require @xmath35 to be non - periodic .
indeed , the family of arcs @xmath218 might be finite , even if @xmath35 is of infinite order .
this occurs typically when @xmath35 is reducible and @xmath10 is contained in a periodic reducible piece of @xmath35 .
however , if we dispose of a homology class @xmath107\in h_1(s,\z)$ ] whose coordinate dual to @xmath10 grows ( i.e. @xmath219 for @xmath220 ) , then the family @xmath221 contains infinitely many distinct arcs since @xmath222 .
[ prop : infinitetrees ] let @xmath7 be a surface obtained by plumbing positive hopf bands according to a tree , other than @xmath0 , @xmath1 , @xmath2 , @xmath3 and @xmath4 .
then , @xmath7 contains infinitely many non - isotopic cutting arcs preserving fibredness .
let @xmath7 be a surface obtained by positive tree - like hopf plumbing .
denote the induced action of the monodromy on @xmath223 by @xmath224 , and let @xmath225 be the basis vectors represented by the core curves of the hopf bands used in the plumbing construction .
it follows from acampo s work on the spectrum of coxeter transformations ( @xcite ) and slalom knots ( @xcite ) , that @xmath224 has a real eigenvalue @xmath226 with @xmath227 if the tree corresponds to neither spherical nor affine coxeter systems .
let @xmath108 be an eigenvector of @xmath224 for the eigenvalue @xmath226 .
then the sequence @xmath228 is unbounded .
choose @xmath229 such that the @xmath230-th coordinate of @xmath108 is nonzero .
let @xmath199 be a spanning arc of the hopf band with core curve @xmath165 .
then @xmath10 is clean , since cutting along @xmath10 yields a connected sum of positive tree - like hopf plumbings , which is fibred .
moreover we have @xmath231 for @xmath220 by construction .
it therefore remains to study the affine coxeter - dynkin trees .
for these , the spectral radius of @xmath224 is equal to one .
however , @xmath224 has a jordan block to the eigenvalue @xmath232 in these cases , and a similar reasoning applies .
combine propositions [ prop : dne7 ] and [ prop : infinitetrees ] .
[ prop:3345 ] the torus links @xmath19 , @xmath20 and @xmath21 admit , up to isotopy ( free on the boundary ) , only a finite number of cutting arcs that preserve fibredness .
[ prop : dne7 ] the positive tree - like hopf plumbings associated to any of the coxeter - dynkin trees @xmath0 , @xmath1 , @xmath2 , @xmath3 or @xmath4 admit , up to isotopy ( free on the boundary ) , only a finite number of cutting arcs that preserve fibredness . before we begin with the proofs , some notation and remarks are necessary .
let @xmath7 be the fibre surface of either @xmath18 , @xmath1 or @xmath3 , and let @xmath39 be the monodromy .
precisely as in section [ sec : monodromies ] , we decompose @xmath7 into finitely many disjoint polygonal disks @xmath70 ( and @xmath71 in the case of torus links ) that are glued using bands ( @xmath72 for the torus links and neighbourhoods of the @xmath156 , @xmath192 for @xmath1 and @xmath3 ) .
we use the letter @xmath141 to denote any of the disks and the letter @xmath200 to denote any of the bands .
let @xmath233 be the union of all the disks , and let @xmath234 be the neighbourhood of the graph on which @xmath35 is assumed to be periodic . 1
. the endpoints of @xmath10 lie in @xmath235 .
[ def : normalpos : b ] for every band @xmath200 , @xmath236 consists of finitely many straight segments parallel to the edges of the graph .
[ def : normalpos : c ] the number of such segments in @xmath200 is minimal among all arcs isotopic to @xmath10 .
4 . @xmath10
intersects the graph transversely in finitely many points of @xmath233 .
5 . @xmath237 , that is , before @xmath10 enters @xmath238 and after it leaves @xmath238 , it stays in the disks that contain its endpoints .
[ def : normalpos : f ] @xmath239 consists of finitely many straight arcs . *
any arc can be brought into normal position by a free isotopy . * if @xmath10 is in normal position , then @xmath41 can be brought into normal position keeping @xmath238 fixed . indeed , it suffices to straighten the two subarcs @xmath240 ( or , undoing the twisting that occurs in the respective annuli ) , sliding the endpoints of @xmath41 along @xmath25 , see figure [ fig : slide ] . + in normal position , keeping @xmath238 fixed.[fig :
slide ] ] * if @xmath10 and @xmath41 are in normal position as above , we may isotope @xmath41 with endpoints fixed and keeping it in normal position , such that @xmath10 and @xmath41 intersect transversely in finitely many points of @xmath233 .
in particular , the sets @xmath241 and @xmath242 are now disjoint ( cf . figure [ fig : transverse ] ) .
+ intersect transversely , keeping normal position.[fig : transverse ] ] * let @xmath10 be in normal position and suppose it passes through at least one band .
let @xmath200 be the first ( respectively last ) band traversed by @xmath10 after ( before ) it starts ( ends ) at a boundary point @xmath145 of one of the disks , say @xmath141 .
then , @xmath145 can not lie between @xmath200 and one of the two bands adjacent to @xmath200 on @xmath142 . otherwise , an isotopy sliding the starting point ( endpoint ) of @xmath10 along @xmath243 would decrease the number of segments in @xmath200 , contradicting part ( [ def : normalpos : c ] ) of definition [ def : normalpos ] .
[ rem : bigons ] suppose @xmath10 intersects @xmath41 .
if @xmath10 is clean , there must be a bigon @xmath244 whose sides consist of a subarc of @xmath10 and a subarc of @xmath41 .
if @xmath245 are in normal position , such @xmath246 takes a particularly simple form : * @xmath246 can not be contained in @xmath233 ( i.e. , in one of the disks @xmath70 or @xmath71 ) .
this would contradict part ( [ def : normalpos : f ] ) of the above definition [ def : normalpos ] .
* none of the two sides of @xmath246 is contained in @xmath233 , since the other side of @xmath246 would have to leave @xmath233 through one of the bands @xmath200 and return through the same @xmath200 .
the disk @xmath246 would then yield an isotopy reducing the number of segments of @xmath247 or @xmath248 , contradicting part ( [ def : normalpos : c ] ) of definition [ def : normalpos ] . * for every band @xmath200 , @xmath249 consists of rectangles with two opposite sides parallel to the edge passing through @xmath200 . *
@xmath250 constists of topological disks @xmath251 connected to at least one rectangle . *
construct a spine @xmath22 for @xmath246 as follows : put a vertex for each @xmath251 and connect two vertices by an edge if the corresponding disks @xmath251 connect to the same rectangle .
@xmath22 is a tree , for @xmath246 is contractible .
two of its vertices correspond to the vertices of the bigon @xmath246 . among the other vertices of @xmath22 ,
there is none of degree one because the adjacent edge would correspond to a rectangle in some @xmath200 whose sides parallel to its core edge both belong to the same arc ( @xmath10 or @xmath41 ) . in other words
, either @xmath10 or @xmath41 would pass through @xmath200 and immediately return through @xmath200 in the opposite direction .
this would contradict part ( [ def : normalpos : c ] ) of definition [ def : normalpos ] .
therefore , @xmath22 is a line consisting of some number of consecutive edges , and the two extremal vertices correspond to the vertices of @xmath246 . [
lem : conflict ] let @xmath245 be in normal position and suppose they intersect in a point @xmath252 , where @xmath141 is one of the disks @xmath70 ( or @xmath71 in the torus link case ) .
let @xmath253 be the components of @xmath254 containing @xmath145 .
if no two of the four points @xmath255 lie in the same band @xmath200 , then @xmath10 can not be clean .
if @xmath10 were clean , there would be a bigon .
after possibly removing a certain number of such bigons , we are left with a bigon @xmath246 with vertex @xmath145 . by remark [ rem : bigons ] , @xmath246 has to leave @xmath141 through one of the adjacent bands .
since one of the sides of @xmath246 is a subarc of @xmath10 and the other side is a subarc of @xmath41 , we find two points among @xmath256 that lie in this band , contradicting the assumption on @xmath253 . [ lem : generalisedconflict ] let @xmath245 be in normal position and let @xmath253 be subarcs of @xmath245 respectively ( not necessarily contained in @xmath233 ) .
suppose that the four endpoints of @xmath257 and @xmath258 are contained in @xmath235 and that no two of them lie on the same band @xmath200 .
we further assume that @xmath257 and @xmath258 intersect in exactly one point and that @xmath253 run through the same bands ( see figure [ fig : generalisedconflict ] )
. then @xmath10 can not be clean .
assume @xmath259 , then @xmath260 .
as in the proof of lemma [ lem : conflict ] , study a bigon @xmath246 that starts at @xmath145 .
@xmath246 consists of a sequence of rectangles as described in remarks [ rem : bigons ] .
starting at @xmath145 , it therefore has to pass through the same bands as @xmath257 and @xmath258 . since @xmath145 was the only intersection between @xmath257 and @xmath258
, @xmath246 has to pass through at least one more band .
but this is impossible by the assumption on the endpoints of @xmath257 and @xmath258 . here
, a sequence of bands @xmath261 is _ consecutive _ , if the set @xmath262 has a connected component that intersects all bands @xmath263 of the sequence in this order , i.e. it is possible to stay on the same side of the graph when walking along the bands .
the twist length @xmath55 denotes the number of edges of @xmath54 enclosed between @xmath264 and @xmath265 , where @xmath264 is a spanning arc of the corresponding boundary annulus that ends at a vertex of @xmath54 ( compare section [ sec : monodromies ] ) .
suppose that @xmath10 is a clean arc in normal position that traverses @xmath57 consecutive bands , @xmath266 .
we may assume that @xmath57 is the maximal number of consecutively traversed bands .
in these bands as well as the adjacent disks , isotope @xmath10 such that it stays on one side of the graph , keeping it in normal position . now bring @xmath41 into normal position transverse to @xmath10 as described in the remarks [ rem : normalpos ] .
recall the description of the monodromy @xmath35 as a twist from section [ sec : monodromies ] : cutting the surface @xmath7 open along the graph results in @xmath267 annuli , where @xmath267 is the number of components of @xmath268 and each annulus has a link component as one boundary and a cycle consisting of edges of the graph as the other boundary . in one of these annuli we will see a subarc @xmath269 that has exactly its endpoints in common with the graph and that travels near the edge boundary for a distance of @xmath57 consecutive edges .
( note that @xmath257 can not have any endpoint on @xmath25 .
this would contradict part ( [ def : normalpos : c ] ) of definition [ def : normalpos ] ) .
let @xmath270 be the disk bounded by @xmath257 and the graph .
the monodromy @xmath35 keeps the link - boundary of this annulus fixed and rotates the neighbourhood @xmath238 of the graph boundary by @xmath55 edges .
since @xmath266 , @xmath271 has one of its endpoints in @xmath270 and the other outside of @xmath270 , so @xmath257 has to intersect its image @xmath271 in a point @xmath260 , and we may assume that @xmath145 is the only intersection between @xmath257 and @xmath271 .
denote by @xmath272 the endpoint of @xmath271 that lies in @xmath270 and let @xmath141 be the disk @xmath70 or @xmath71 containing @xmath272 . then make sure that @xmath252 by an isotopy on @xmath271 preserving normal position if necessary ( compare figures [ fig : annulus ] and [ fig : crossing ] ) .
however @xmath10 is clean , so there must be a bigon in @xmath7 whose sides consist of a subarc of @xmath10 and a subarc of @xmath41 .
after possibly removing a certain number of such bigons , we will be left with a bigon @xmath246 starting at @xmath145 . from the remarks
[ rem : bigons ] we know that @xmath246 has to leave @xmath141 and consists of a sequence of rectangles .
let @xmath273 be the first rectangle in this sequence , i.e. @xmath273 is contained in a band adjacent to @xmath141 .
let @xmath274 be the two bands adjacent to @xmath141 that contain segments of @xmath257 , @xmath275 being the one that also contains a segment of @xmath271 ( see figure [ fig : crossing ] ) .
let @xmath28 be the component of @xmath276 that contains @xmath272 .
we claim that @xmath28 can not leave @xmath141 through @xmath277 nor @xmath275 . indeed ,
if @xmath28 would leave @xmath141 through @xmath277 , @xmath41 would traverse @xmath278 consecutive bands , contradicting the assumption on @xmath57 being maximal . on the other hand ,
if @xmath28 would leave @xmath141 through @xmath275 , we could reduce the number of segments in @xmath279 , contradicting the normal position of @xmath41 , i.e. part ( [ def : normalpos : c ] ) of definition [ def : normalpos ] .
in contrast , @xmath10 leaves @xmath141 , starting from @xmath145 in both directions , through @xmath277 and @xmath275 .
consider now the subarcs of @xmath10 and @xmath41 that constitute two opposite sides of the rectangle @xmath273 .
since @xmath273 is contained in a band adjacent to @xmath141 , these two subarcs arrive at @xmath142 through the same band , and they connect directly to @xmath252 .
therefore , we must have @xmath280 , since @xmath275 is the only band containing two subarcs of @xmath10 and @xmath41 that directly connect to @xmath252 .
furthermore , @xmath273 has to be the region enclosed by @xmath281 and @xmath282 .
following @xmath271 in the direction from @xmath272 to @xmath145 , we see that it leaves @xmath141 through @xmath275 as one of the sides of @xmath273 and continues staying on the same side of the graph for exactly @xmath183 more edges . by assumption
, @xmath145 is the only intersection between @xmath257 and @xmath271 , so the bigon @xmath246 has to continue for at least @xmath183 more rectangles through consecutive bands .
similarly , the sides of these rectangles that are subarcs of @xmath10 have to continue for at least @xmath183 consecutive bands .
we obtain a contradiction to the maximality of @xmath57 , because @xmath257 ends after @xmath283 bands starting from @xmath141 , since @xmath35 rotates the graph by @xmath55 edges .
this finishes the proof .
we will concentrate on the most complicated case of the torus knot @xmath21 .
it contains all difficulties appearing in the proofs for @xmath19 and @xmath20 which go along the same lines with fewer cases to consider . for each link appearing in table [ table ] of section [ sec : exceptional ]
, we will indicate one ( but not every ) possible choice of a cutting arc that yields the link in question .
let hence @xmath7 be the fibre surface of @xmath21 and let @xmath199 be any arc that preserves fibredness , i.e. a clean arc .
bring @xmath10 into normal position using an isotopy ( not fixing the boundary ) , cf .
remarks [ rem : normalpos ] .
since @xmath35 permutes the vertices @xmath284 cyclically as well as the vertices @xmath285 , it suffices to show that there are only finitely many clean arcs starting at a point of @xmath286 or at a point of @xmath287 , up to isotopy .
we may further assume that @xmath10 starts either at a point of @xmath286 between @xmath127 and @xmath288 or at a point of @xmath287 between @xmath289 and @xmath119 . _
_ @xmath10 starts at @xmath286 , between @xmath127 and @xmath288 .
then , @xmath10 can not continue through either of the bands @xmath290 nor @xmath291 by the last item of remarks [ rem : normalpos ] .
so , either @xmath10 stays in @xmath176 ( and there are only four such arcs up to isotopy ) , or it continues through @xmath292 or @xmath293 .
if @xmath10 stays in @xmath176 , the links obtained by cutting are @xmath3 ( e.g. if @xmath10 ends between @xmath127 and @xmath294 ) and @xmath295 ( e.g. if @xmath10 ends between @xmath294 and @xmath126 ) . _
_ @xmath10 continues through @xmath296 . arriving in @xmath297 , there are three possibilities : either @xmath10 ends at a point of @xmath298 between @xmath129 and @xmath299 ( and cutting along @xmath10 yields @xmath300 ) , or it continues through @xmath301 or @xmath302 ( ending at other points of @xmath298 is impossible by the last item of remarks [ rem : normalpos ] ) .
_ _ @xmath10 continues through @xmath301 . arriving in @xmath303
, @xmath10 can end at a point of @xmath128 ( cutting yields @xmath304 if @xmath10 ends between @xmath305 and @xmath306 , and @xmath307 summed with the unknot component of @xmath215 if @xmath10 ends between @xmath125 and @xmath305 ) , or it can continue through @xmath308 or @xmath309
. it can not continue through @xmath310 , since @xmath311 is a sequence of three consecutive bands , so @xmath10 would not be clean by lemma [ lem : consecutive ] .
finally , @xmath10 can not continue through @xmath312 .
if it did , @xmath10 and @xmath41 would intersect in a point of @xmath176 , and lemma [ lem : conflict ] would imply that @xmath10 can not be clean ( see figure [ fig : t35 ] , top left ) . note that we do not know whether the mentioned intersection is the only one since we do not know how @xmath10 ends . _
_ @xmath10 continues through @xmath308 . from @xmath313
, it can not continue through @xmath314 , for @xmath315 are consecutive ( lemma [ lem : consecutive ] ) . if it continues through @xmath316 it can not continue through any band adjacent to @xmath155 . indeed , @xmath317 are consecutive , so @xmath10 can not continue through @xmath302 . if it would continue through @xmath318 or @xmath319 or @xmath320 , we could apply lemma [ lem : generalisedconflict ] to the band @xmath316 to show that @xmath10 is not clean ( see figure [ fig : t35 ] ) . _
_ @xmath10 continues through @xmath309 .
if it ends in @xmath321 between @xmath322 and @xmath323 , we obtain @xmath324 after cutting .
otherwise , it can continue from @xmath321 through @xmath293 or through @xmath318 .
_ _ if it continues through @xmath293 , it can not go further .
firstly , @xmath325 are consecutive , so @xmath291 is no option ( lemma [ lem : consecutive ] ) .
neither can it proceed through @xmath290 ( this would produce a self - intersection of @xmath10 ) nor @xmath296 ( for otherwise we could apply lemma [ lem : conflict ] to an intersection between @xmath10 and @xmath41 in @xmath176 ) .
if it continues through @xmath314 , it can not go on through @xmath308 since @xmath326 are consecutive ( lemma [ lem : consecutive ] ) .
suppose it continues through @xmath316 . from @xmath155
, it can not proceed through any of @xmath327 , for otherwise we could apply lemma [ lem : generalisedconflict ] to the bands @xmath314 and @xmath316 , with an intersection between @xmath10 and @xmath41 occuring in @xmath155 ( see figure [ fig : t35 ] left ) .
however , @xmath10 can not continue through @xmath302 either , because we could again apply lemma [ lem : generalisedconflict ] , this time for the band @xmath309 and an intersection in @xmath303 ( see figure [ fig : t35 ] right ) . _
_ @xmath10 continues from @xmath321 through @xmath318 .
if it ends in @xmath155 between @xmath328 and @xmath119 , cutting yields @xmath329 . otherwise , it can not continue from @xmath155 through @xmath316 since @xmath330 are consecutive . neither
can it proceed through @xmath302 ( apply lemma [ lem : conflict ] to @xmath155 ) .
so @xmath10 can only continue through @xmath319 or @xmath320 .
_ _ if it continues through @xmath319 , the only option to go further is through @xmath291 , since @xmath331 are consecutive . from @xmath176 ( compare figure [ fig : t35 ] ) , it can not continue through @xmath290 nor @xmath296 ( apply lemma [ lem : generalisedconflict ] to @xmath291 with an intersection occuring in @xmath176 ) .
neither can it continue through @xmath293 , since @xmath332 are consecutive .
so it has to go through @xmath314 . arriving in @xmath313
, it can not continue through @xmath308 ( apply lemma [ lem : generalisedconflict ] to @xmath318 with an intersection occuring in @xmath321 ) .
therefore @xmath10 has to continue through @xmath316 . from @xmath155
, it can not proceed further .
firstly , @xmath302 is not an option ( otherwise apply lemma [ lem : generalisedconflict ] to @xmath318 and @xmath309 with an intersection in @xmath303 ) .
neither can it go through @xmath318 or @xmath319 ( apply lemma [ lem : generalisedconflict ] to @xmath333 with an intersection occuring in @xmath155 ) .
finally , it can not pass through @xmath320 either ( apply lemma [ lem : generalisedconflict ] to the bands @xmath333 with an intersection occuring in @xmath155 ) .
_ _ if it continues through @xmath320 and arrives in @xmath334 , it can not proceed through @xmath290 ( apply lemma [ lem : generalisedconflict ] to @xmath301 with an intersection occuring in @xmath297 , see figure [ fig : t35 ] left ) .
so it has to go through @xmath310 . from @xmath303
, it can not proceed through @xmath301 , for @xmath335 are consecutive . neither
can it go through either of @xmath308 nor @xmath309 ( apply lemma [ lem : conflict ] to an intersection occuring in @xmath303 , see figure [ fig : t35 ] right ) .
finally , @xmath312 can be ruled out by lemma [ lem : generalisedconflict ] , applied to the bands @xmath301 and @xmath296 , with an intersection occuring in @xmath176 . _
_ @xmath10 continues through @xmath302 ( see figure [ fig : t35 ] ) . arriving in @xmath155
, it can not continue through any band .
firstly , @xmath336 are consecutive , so @xmath10 can not continue through @xmath316 . if it would continue through any of the other bands adjacent to @xmath155 , @xmath10 would intersect @xmath41 in a point of @xmath155 such that we could apply lemma [ lem : conflict ] to obtain a contradiction to @xmath10 being clean . _ _ @xmath10 proceeds through @xmath314 .
if it ends in @xmath313 between @xmath125 and @xmath337 , we obtain @xmath338 after cutting . from @xmath313
, it can continue through @xmath308 or through @xmath316 . _
_ @xmath10 continues through @xmath308 .
it can not go on via @xmath301 , for @xmath339 are consecutive .
neither can it continue through @xmath310 or @xmath312 by lemma [ lem : conflict ] applied to an intersection in @xmath176 .
if it next passes through @xmath309 , it can not go on through @xmath293 , because @xmath340 are consecutive . proceeding through @xmath318
, it can end in @xmath155 between @xmath119 and @xmath299 ( this yields @xmath341 ) .
however , the only possibility for @xmath10 to go further is via @xmath302 , for @xmath330 are consecutive ( so @xmath316 is no option ) , and @xmath10 can not continue through @xmath319 nor @xmath320 by applying lemma [ lem : generalisedconflict ] to the band @xmath318 with an intersection of @xmath245 in @xmath155 .
so @xmath10 continues through @xmath302 and arrives in @xmath297 .
from there , it can not continue through @xmath296 ( apply lemma [ lem : generalisedconflict ] to @xmath301 and an intersection in @xmath313 ) .
if it continues through @xmath301 , it can not go further : @xmath308 is impossible because @xmath342 are consecutive , @xmath309 can be excluded by lemma [ lem : conflict ] , applied to @xmath303 , and @xmath310 as well as @xmath312 can be ruled out by lemma [ lem : generalisedconflict ] , applied to @xmath308 and @xmath314 with an intersection occuring in @xmath176 .
_ _ @xmath10 continues through @xmath293 .
this is analogous to case a.1 .
+ _ _ @xmath10 starts at @xmath287 between @xmath289 and @xmath119 .
then , it can only continue through @xmath290 by the last item of remarks [ rem : normalpos ] . from @xmath176
, it can proceed through four distinct bands . _
_ @xmath10 continues through @xmath291 .
since @xmath343 are consecutive , it can a priori only continue through @xmath319 .
but this is impossible as well by lemma [ lem : conflict ] , applied to the intersection between @xmath10 and @xmath41 occuring in @xmath334 . _
_ @xmath10 continues through @xmath309 . from @xmath303
, it can not continue through @xmath308 because @xmath344 are consecutive ( lemma [ lem : consecutive ] ) .
neither can it go on through @xmath301 nor @xmath310 ( apply lemma [ lem : conflict ] to @xmath176 ) .
suppose @xmath10 continues through @xmath312 . from @xmath345
, it can not go on via @xmath319 since @xmath346 are consecutive .
if it proceeds via @xmath319 , we can apply lemma [ lem : generalisedconflict ] to the band @xmath290 with an intersection in @xmath334 to obtain a contradiction . _
_ @xmath10 continues through @xmath318 . from @xmath155 , there are only two options for @xmath10 to proceed further .
indeed , @xmath347 are consecutive , so @xmath319 is out of the question .
@xmath320 can be ruled out by lemma [ lem : conflict ] for @xmath155 .
the remaining possibilities are @xmath302 and @xmath316 . _
_ @xmath10 continues through @xmath302 .
from there , it can not continue through @xmath301 ( apply lemma [ lem : generalisedconflict ] to @xmath302 ) .
so it has to branch off via @xmath296 to @xmath176 .
from there , it can not continue through @xmath291 since otherwise @xmath10 would self intersect in @xmath176 .
@xmath290 is impossible as well , for @xmath348 are consecutive .
@xmath291 can be ruled out using lemma [ lem : conflict ] for @xmath155 .
so @xmath10 can only continue through @xmath314 , and from there only through @xmath308 ( @xmath349 are consecutive ) . from @xmath303 , it can not go on through any band except @xmath312 .
indeed , @xmath301 is impossible because @xmath339 are consecutive .
@xmath310 and @xmath309 can be ruled out by applying lemma [ lem : generalisedconflict ] to @xmath350 and @xmath308 respectively . after passing through @xmath312
, @xmath10 can not go further : @xmath291 is impossible by lemma [ lem : generalisedconflict ] ( applied to @xmath351 ) and @xmath319 can be ruled out by applying lemma [ lem : generalisedconflict ] to @xmath352 . _ _ @xmath10 continues through @xmath316 .
then , @xmath314 can not be next since @xmath353 are consecutive .
thus @xmath10 passes through @xmath308 . from @xmath303
, it can not go on via @xmath309 , for @xmath354 are consecutive .
@xmath310 and @xmath301 are impossible as well ( apply lemma [ lem : generalisedconflict ] to @xmath293 ) .
so @xmath10 has to go through @xmath312 .
then , it can not proceed through @xmath291 ( apply lemma [ lem : generalisedconflict ] to @xmath312 ) .
it can not go via @xmath319 either ( apply lemma [ lem : generalisedconflict ] to @xmath355 ) , so @xmath10 can not continue at all .
we will present a case by case analysis for the possible clean arcs @xmath10 in the fibre surface @xmath7 of each of @xmath3 and @xmath1 .
the reader interested in studying the proof is advised to follow the arguments along with a pencil and copies of figures [ fig : e7hexagons ] and [ fig : dnsquares ] , top and bottom .
as in the proof of proposition [ prop:3345 ] above , we will make extensive use of lemma [ lem : consecutive ] to exclude further polygon edges that @xmath10 might cross on its way from its starting point to its end . in order to keep the proof short
, we will usually refer to such situations by just saying `` @xmath10 is trapped '' , or by saying that an edge `` is a trap '' , meaning that @xmath10 would traverse too many consecutive bands to be clean .
+ * ( @xmath3)*first , let @xmath7 be the fibre surface of @xmath3 , denote its monodromy @xmath35 and let @xmath199 be a clean arc . bring @xmath10 into normal position with respect to @xmath356 . note that the set of vertices of the hexagons @xmath167 decompose into two orbits under @xmath35 , namely the orbit of the vertex of @xmath176 between @xmath160 and @xmath357 , and the orbit of the vertex of @xmath176 between @xmath357 and @xmath358 .
we may therefore assume by remark [ rem : clean ] that @xmath10 starts at one of these two vertices . _
_ @xmath10 starts at the vertex of @xmath176 between @xmath160 and @xmath357 .
define an involution @xmath45 as follows : @xmath359 interchanges hexagons @xmath176 and @xmath303 and then reflects @xmath167 along the diagonals parallel to @xmath358 , @xmath360 , @xmath160 respectively , whereby it induces the permutation @xmath361 on the edges @xmath362 .
we have @xmath46 , @xmath363 fixes the vertex of @xmath176 between @xmath160 and @xmath357 and swaps the edges @xmath364 as well as the edges @xmath365 . by remark [ rem :
clean ] , we may therefore assume that @xmath10 either continues through @xmath366 or through @xmath367 . _
_ @xmath10 continues through @xmath366 . from @xmath155
, it can only choose @xmath368 .
indeed , @xmath358 , @xmath369 and @xmath160 would imply an intersection in @xmath176 ( lemma [ lem : conflict ] , compare figure [ fig : e7case11 ] , left ) , and @xmath370 is consecutive to @xmath366 ( lemma [ lem : consecutive ] , applied to the component with twist length one ) .
arriving in @xmath303 , @xmath357 and @xmath369 would imply an intersection in @xmath303 , so continuation is possible through @xmath370 , @xmath367 , @xmath360 only .
but if @xmath10 continues through @xmath367 or @xmath360 , it will be trapped ( compare figure [ fig : e7case11 ] , right ) .
therefore it goes through @xmath370 . arriving in @xmath155
, it has to go through @xmath366 ( @xmath368 implies an intersection in @xmath155 and @xmath160 , @xmath369 , @xmath358 imply intersections in @xmath176 ) .
however , passing through @xmath366 , @xmath10 is trapped . _
_ @xmath10 continues through @xmath367 . from @xmath303
, it can continue through @xmath370 , @xmath369 , @xmath360 or @xmath368 ( @xmath357 implies an intersection in @xmath303 ) .
if it passes through @xmath369 or @xmath368 , it is trapped .
so @xmath360 and @xmath370 are the only possiblities left . _
_ @xmath10 continues through @xmath360 . upon arrival in @xmath176
, it can not continue through @xmath357 , @xmath367 ( intersection in @xmath303 ) nor through @xmath160 ( this would imply an intersection in @xmath176 ) .
but if it continues through either of @xmath358 or @xmath366 , it is trapped . _
_ @xmath10 continues through @xmath370 . from @xmath155
, @xmath10 can not go on through @xmath368 , @xmath160 ( this would imply an intersection in @xmath155 ) .
if it passes through @xmath366 , it is trapped .
suppose it continues through @xmath369 .
arriving in @xmath303 , it can not continue through @xmath368 , @xmath360 , @xmath370 ( this would produce an intersection in @xmath155 ) , nor through @xmath357 ( intersection in @xmath303 ) . finally , continuing through @xmath367 , it will be trapped .
therefore @xmath10 has to continue from @xmath155 through @xmath358 . arriving in @xmath176
, it can continue through @xmath357 , @xmath360 or @xmath366 ( @xmath160 implies an intersection in @xmath176 and @xmath367 implies an intersection in @xmath303 ) . but all of these are traps . _
_ @xmath10 starts at the vertex of @xmath176 between @xmath357 and @xmath358 .
define an involution @xmath371 as follows : @xmath372 interchanges @xmath176 and @xmath303 and then reflects @xmath167 along the diagonals parallel to @xmath160 , @xmath357 , @xmath366 respectively , inducing the permutation @xmath373 on the edges . as in case 1 , we have @xmath374 , and @xmath375 fixes the vertex of @xmath176 between @xmath357 and @xmath358 , swapping @xmath366 and @xmath367 as well as @xmath160 and @xmath360 . by remark [ rem :
clean ] , we may therefore assume that @xmath10 continues through either @xmath160 or @xmath367 .
however , if @xmath10 continues through @xmath160 , it is trapped .
therefore it continues through @xmath367 . from @xmath303
, it can continue through @xmath360 or @xmath368 ( @xmath369 is a trap and @xmath357 , @xmath370 imply intersections in @xmath303 ) . _
_ @xmath10 continues through @xmath360 . from @xmath176
, it can not continue through any of @xmath357 , @xmath160 , @xmath367 , because this would produce an intersection in @xmath303 , and @xmath358 is a trap .
therefore , it continues through @xmath366 and arrives in @xmath155 .
continuation through @xmath160 produces an intersection in @xmath155 , and @xmath369 , @xmath358 , @xmath370 imply intersections in @xmath176 .
finally , @xmath368 is a trap . _
_ @xmath10 continues through @xmath368 . arriving in @xmath155
, it can only continue through @xmath160 or @xmath366 ( any other continuation produces an intersection in @xmath176 ) . however , both @xmath160 and @xmath366 are traps , ending the proof for @xmath3 .
+ * ( @xmath1 , @xmath57 even)*now , suppose @xmath57 is even and let @xmath10 be a clean arc in the fibre surface @xmath7 of @xmath1 in normal position with respect to @xmath376 .
define an involution @xmath45 as follows : @xmath359 permutes the disks @xmath70 according to the rule @xmath377 for @xmath378 and then reflects every @xmath70 on the diagonal that contains the vertex between @xmath156 and @xmath379 ( all indices are to be taken modulo @xmath57 ) . again
@xmath46 , and @xmath363 fixes the vertex of @xmath176 between @xmath380 and @xmath381 as well as the vertex between @xmath160 and @xmath370 , and swaps the other two vertices .
we may therefore assume that @xmath10 starts at a vertex of @xmath176 which is not the vertex between @xmath381 and @xmath370 . _
_ @xmath10 starts at the vertex of @xmath176 between @xmath160 and @xmath380 .
if it continues through @xmath381 , it is already trapped .
so it has to continue through @xmath370 . arriving in @xmath155
, it can continue through @xmath382 , @xmath383 or @xmath367 . _
_ @xmath10 continues from @xmath155 through @xmath382 . from @xmath303
, it can not continue through @xmath357 ( otherwise it would intersect with @xmath41 ) , so it can only proceed through @xmath381 or @xmath366 . however , both are traps .
_ _ @xmath10 continues from @xmath155 through @xmath383 .
this is similar to case 1.1 : arriving in @xmath384 , @xmath10 can only continue through @xmath385 ( which is a trap ) or @xmath366 .
if it goes through @xmath366 , it has to continue from @xmath303 through @xmath382 ( @xmath357 produces an intersection in @xmath303 and @xmath381 produces an intersection in @xmath155 ) .
then however , it is trapped again . _
_ @xmath10 can therefore continue from @xmath155 through @xmath367 only . in @xmath386 ,
the same situation reproduces , except that all indices in consideration are now shifted by @xmath387 .
therefore the only way for @xmath10 to continue from @xmath386 is by passing through the edges @xmath388 after at most @xmath389 more steps , @xmath10 will be trapped . _
_ @xmath10 starts at the vertex of @xmath176 between @xmath380 and @xmath381 . using @xmath359 again
, we may assume that it continues through @xmath160 to @xmath390 .
if it goes through @xmath391 next , it is trapped since it is forced to follow the sequence of edges @xmath392 if it goes through @xmath393 to @xmath394 instead , it can only continue from there through @xmath395 or @xmath396 , and these are traps again .
so it has to continue from @xmath390 through @xmath397 . in @xmath398 , the same situation as one step earlier ( where @xmath10 arrived through @xmath160 in @xmath390 ) reproduces , except that all indices appearing in the consideration are now shifted by @xmath399 .
hence the only way @xmath10 can continue from @xmath398 is by going through the sequence of edges @xmath400 after at most @xmath389 steps , @xmath10 will be trapped . _
_ @xmath10 starts at the vertex of @xmath176 between @xmath160 and @xmath370 . using the involution @xmath359 from above
, we may assume that it continues through @xmath381 . from @xmath303 , it can not go on through @xmath366 , for this would imply an intersection in @xmath303 . however , the two possibilities that remain ( @xmath382 and @xmath357 ) are traps , which ends the proof for @xmath1 , @xmath57 even .
+ * ( @xmath1 , @xmath57 odd)*finally , let @xmath57 be odd and let @xmath7 be the fibre surface of @xmath1 .
suppose again we have a clean arc @xmath199 in normal position with respect to @xmath401 .
since the monodromy permutes the @xmath70 cyclically and since there are only two orbits of vertices of the @xmath70 , we may assume that @xmath10 starts in @xmath176 , at the vertex between @xmath160 and @xmath357 , or at the vertex between @xmath357 and @xmath402 .
as before , we then make use of remark [ rem : clean ] with the help of the involution @xmath45 defined as follows : @xmath377 by translations followed by a reflection on the diagonal of @xmath70 that contains the vertex between @xmath403 and @xmath404 for @xmath405 and reflection on the diagonal of @xmath70 that contains the vertex between @xmath156 and @xmath172 for @xmath406 . applying remark [ rem : clean ] as before , we may assume that @xmath10 either starts at the vertex of @xmath176 between @xmath407 and continues through @xmath402 ( say ) , or that it starts at the vertex of @xmath176 between @xmath357 and @xmath402 , continuing through @xmath160 ( say ) .
so there are two cases to consider , one being very similar to case 1 above and the other similar to case 3 .
no new arguments are needed . | among all torus links , we characterise those arising as links of simple plane curve singularities by the property that their fibre surfaces admit only a finite number of cutting arcs that preserve fibredness .
the same property allows a characterisation of coxeter - dynkin trees ( i.e. , @xmath0 , @xmath1 , @xmath2 , @xmath3 and @xmath4 ) among all positive tree - like hopf plumbings . |
in 1988 haxton@xcite proposed @xmath0i as the active ingredient in a new solar - neutrino detector .
he estimated that a tank containing 1000 tons of iodine would detect about 20 times as many neutrinos as the chlorine experiment at homestake .
he also noted that a 3/2@xmath5 state at 125 kev in @xmath0xe is accessible to neutrinos from @xmath4be , and that if the ratio of @xmath4be and @xmath2b cross sections were substantially different from that in chlorine , then the two experiments combined could determine the fluxes of both the @xmath4be and @xmath2b neutrinos .
since the @xmath4be flux now appears to be a critical piece of the solar - neutrino puzzle@xcite , a high - statistics iodine detector has become attractive . in 1991
we attempted@xcite to improve on haxton s estimate by using the quasiparticle tamm - dancoff approximation ( qtda ) to calculate the response of iodine to both @xmath2b and @xmath4be neutrinos .
we concluded that haxton s estimate was a little too high , but our results were still sufficiently encouraging ( and carried enough uncertainty ) to make a direct calibration desirable .
in fact , around the same time an attempt was made@xcite to measure the gamow - teller ( gt ) strength distribution , which determines solar - neutrino cross sections , via the charge - exchange reaction @xmath0i(_p , n_)@xmath0xe .
unfortunately , in cases for which no states in the final nucleus are accessible by beta decay , the reaction determines the gt distribution only up to an overall normalizing constant . the solar neutrino cross section in @xmath0i
can therefore not be extracted without measuring at least one more quantity for example , the absolute gt strength to a particular state , or the total integrated strength below a certain energy . to fix the normalization of the gt distribution , and to demonstrate that the counting of @xmath0xe works
as expected , a group working at lampf has recently exposed an iodine target to a known flux of electron neutrinos from muon decay@xcite .
the underlying idea is to use the total lampf cross section to fix the unknown constant multiplying the gt distribution .
this task , unfortunately , is not completely straightforward . because the lampf neutrinos have on average an energy of about 30 mev ,
their wavelengths are short enough to change the form of the cross section ; forbidden operators that depend on nucleon coordinates and momenta can modify the `` allowed '' strength from the gt operator @xmath6 . for the experiment to yield a reliable normalization , the forbidden contributions must either be small or accurately calculated and removed . the current state of affairs has prompted us to reexamine neutrino absorption by @xmath0i . here ,
after reviewing our prior work in sec [ s : prior ] and developing an improved version in sec.[s : new ] , we calculate the lampf cross section , including the forbidden corrections .
these turn out to be large enough to complicate an accurate extraction of the allowed component from experimental data .
in ref . @xcite we developed an approximation for odd - mass nuclei that is closely related to the usual even - even qtda .
the starting point is the assumption that the ground state of @xmath0i is predominantly of one - quasiparticle character , @xmath7 where @xmath8 is the fully - paired bcs quasiparticle vacuum and @xmath9 creates a proton quasiparticle in the @xmath10 orbit .
we confirmed our assumption in part by adding three - quasiparticle states of the form @xmath11 ^{5/2 } \left| bcs \right\rangle , \ ] ] the admixtures of which turned out to be small .
( in the above equation _ i , k , l _ are valence orbitals and @xmath12 creates a quasineutron in orbit @xmath13 . ) the form of the iodine ground state implies that the space of states in xenon accessible via neutrino absorption is largely spanned by the analogous set @xmath14 ^j
\left| bcs \right\rangle , \ ] ] where @xmath15 and @xmath16 assume the values @xmath17 , and @xmath18 , and @xmath19 is any intermediate angular momentum .
we therefore used the set eq .
[ e : xestates ] in calculating the @xmath0xe spectrum and the gt strength distribution .
our valence space for both protons and neutrons consisted of the @xmath20 oscillator shell plus the two @xmath21 orbitals from the next oscillator shell .
we took the two - body interaction from ref .
@xcite and modified it by scaling the pairing matrix elements to reproduce empirical pairing gaps and replacing the neutron - proton monopole - monopole component with a constant average interaction .
we changed the monopole component because it alters single - particle energies through mean field effects , which we accounted for phenomenologically by taking our single - particle energies from a wood - saxon potential with parameters appropriate for @xmath0i .
the end result of our calculation was ( assuming standard - solar - model fluxes ) a @xmath2b cross section of 2.2@xmath3 ( 13 snu ) smaller than haxton s estimate by almost a factor of three and a @xmath4be cross section of @xmath22 ( 9.4 snu ) .
the ratio of @xmath4be to @xmath2b cross sections was much larger than in chlorine .
there were of course drawbacks in our approach .
first of all , the qtda is number - nonconserving ; fluctuations in particle number introduce some error , which we were unable to estimate .
in addition , we omitted basis states that might prove important despite the argument above .
for example , three - quasineutron configurations in @xmath0xe should combine with the states in eq .
[ e : xestates ] to create a low - lying collective @xmath23 phonon , which could in turn modify the strength distribution at low excitation energies . in @xmath0i , a quadrupole phonon formed from three - quasiproton configurations plus the states in eq .
[ e : istates ] could significantly dilute the one - quasiparticle content of the ground state .
furthermore , our two - body interaction was not entirely consistent with our selection of single - particle energies because we retained the like - particle monopole component while arguing that the neutron - proton monopole interaction ought to be excluded .
both can modify single - particle energies and , moreover , neither is trustworthy no matter how the energies are chosen ( for a convincing demonstration see ref.@xcite ) .
finally , we took no account of spreading widths associated with our states in xenon . in revisiting our calculation
we have addressed all these issues .
we have now improved our treatment of neutrino absorption by @xmath0i in several ways . to determine the error introduced by number - nonconservation in the qtda we have recalculated the gt distribution in the number - conserving generalized - seniority ( gs)@xcite approximation ( which is equivalent to qtda in all other respects ) .
the one- and three - quasiparticle configurations above correspond to well - defined states in the gs approach .
the comparison will be discussed in detail shortly .
we have also made several improvements to the effective force .
for example we have now removed the monopole component of the interaction both in the neutron - proton and like - particle channels , and have not replaced them with anything else .
the average interaction that we included earlier has no effect ( except on binding energies ) in the number - conserving gs approach and so can be omitted .
we argued previously that some remnant of the monopole force might be needed in a qtda treatment , where because of the mixing of particle numbers even a constant interaction can change wave functions and energies .
we now find , however , that the effects of number - nonconservation are minimized when the monopole force is dropped altogether .
in addition , we no longer scale the like - particle interaction in the pairing channels . though the scaling improves agreement with pairing gaps it badly affects properties of low - lying states ( e.g. energies and magnetic moments ) .
beside altering the two - body interaction , we have also changed the way we select single - particle energies . in our earlier work
, we used the energies given in ref .
@xcite ; here we take them instead from the spectra of n=81 nuclei ( for neutrons ) and z=51 nuclei ( for protons ) that are close to @xmath0i and @xmath0xe .
the new single - particle energies yield a better ordering of low - lying levels in both mass-127 nuclei the old ones reversed the order , e.g. , of the @xmath24 and @xmath18 levels in @xmath0i .
another important change is the inclusion in the qtda of the three - quasiproton and three - quasineutron configurations discussed above . and
finally , we have now taken some account of the background of even more complex states by giving the final states a spreading width .
below we discuss all of these changes and their effects in detail . in the gs approximation
the bcs vacuum is replaced by a number - conserving involving a condensate of @xmath25 @xmath26-pairs for protons and a corresponding condensate for neutrons , viz : @xmath27 where ( @xmath28 ) @xmath29 ^ 0\ ] ] createsa coherent @xmath25 pair , the state @xmath30 contains no valence particles , @xmath31 and @xmath32 are the number of proton and neutron valence @xmath26 pairs , @xmath33 denotes the generalized seniority for particles of type @xmath34,and the @xmath35 are ( variationally - determined ) structure constants related to the @xmath36 s and @xmath37 s of the bcs formulation . in this framework , our original qtda treatment of xenon corresponds to diagonalizing the shell - model hamiltonian in the space of states @xmath38 ^j \left| n_{p } - 1,n_{n } \right\rangle , \nonumber\end{aligned}\ ] ] where @xmath39 and @xmath40 now create _ real _ particles .
an analogous construction yields the states in iodine corresponding to those in eq .
[ e : istates ] .
( the one spurious state in each nucleus with @xmath41 is easily removed . ) in table [ t : energies ] we display the energies of low - lying states , alongside those calculated in the gs and qtda approximations ( without three - like - quasiparticle states or their gs equivalents ) .
neither calculation does terribly well , but two facts are important .
first , the two agree well with one another .
second , the calculations agree worst with experiment when three - quasiparticle configurations are important .
this problem will partly be remedied later when we add the three - like - quasiparticle states .
[ f : str1 ] shows the gamow - teller strength up to 8 mev the strength above 7.23 mev , the neutron - emission threshold is irrelevant for _ radiochemical _ neutrino detection calculated in the same two schemes . here , as in ref .
@xcite we have reduced the strengths by a phenomenological quenching parameter ( 1.0/1.26)@xmath42 .
the validity of this simple prescription , which in weak interactions corresponds to setting @xmath1 to -1.0 , is not universally accepted and we will return to it later .
for now we note the remarkable agreement between the two methods .
the total strength below threshold is 4.38 in the gs calculation and 4.13 in the qtda .
the @xmath2b cross sections are 4.39@xmath3 and 3.91@xmath3 , respectively , close to one another but also markedly larger than that reported in ref.@xcite ( the largest part of the change is due to the pairing matrix elements , which we no longer renormalize ) .
the only important disagreement between the two calculations is in the strength to the lowest @xmath43 state : .079 in the gs scheme and .042 in the qtda .
though it is unreasonable to expect calculations designed to reproduce the entire spectrum to match up state by state , in this instance the disagreement is disappointing because this one strength completely determines the cross section for @xmath4be neutrinos .
the gs strength is more than twice as large as in our previous paper ( .035 ) and if correct would ( again with standard solar fluxes ) result in more events from @xmath4be than from @xmath2b . by comparison
the qtda result for the first @xmath43 state is only marginally bigger than before .
the gs calculation is clearly preferable since it does not violate particle number , but because of its complexity we have not extended it to include more configurations ( which can also affect the @xmath4be strength ) , as we have in the qtda .
we therefore can not make a terribly strong statement about the @xmath4be cross section ; all signs , however , point to its being much larger with respect to the @xmath2b cross section than in @xmath44cl .
other properties are very similar in the two approaches provided all monopole forces are set to zero . for the rest of this paper , then , we will no longer use the gs scheme , since the qtda is computationally much simpler to extend .
we next examine the effects of the three - quasineutron and three - quasiproton configurations in the qtda .
table [ t : energies1 ] shows the same energy levels as in table [ t : energies ] , but with the extra configurations included .
the result is a dramatic lowering in energy of some of the mainly three - quasiparticle states , reflecting the formation via the neutron - proton interaction of a collective quadrupole phonon .
the new configurations also increase the density of low - lying states substantially . though the quantitative agreement in the lowest states is still not impressive , these new results are a considerable improvement . furthermore ,
as noted above , the method is designed to encompass the entire final - state spectrum ( in our model space several thousand states ) ; the lowest lying states , with the exception of the first @xmath43 , are no more important than the others .
the formation of phonon - like states shifts the e2 strength downwards , although no single multiplet is as collective as implied by experiment
. this may be an indication that further correlations associated with higher quasiparticle number ( or with deformation ) ought to be included at some level .
the same conclusion might be drawn from the @xmath0i ground - state magnetic moment , which changes ( with free g - factors ) from 4.39 to 3.76 .
( the experimental value is 2.81 and the schmidt value is 4.79 . ) the incomplete but still significant improvement implies that spin correlations , important for parts of the gt distribution , are represented better than before , though again not perfectly .
it also suggests that at least some extra quenching of the spin operator is in order . in fig .
[ f : str2 ] we compare the qtda gt strength below the neutron - emission threshold with and without the three - like - quasiparticle configurations .
the difference , while nonnegligible , is not dramatic .
the low - lying phonon - like states apparently do little to the charge - changing strength , which is in a different channel . the main effect , a small overall loss of strength is due largely to the reduction of the single - quasiparticle component of the @xmath0i ground state from 86% to 76% .
the ground state three - quasiparticle components go mainly into _
five_-quasiparticle states in @xmath0xe , which are not included in our basis .
to the extent that these states lie below the neutron - emission threshold ( they should begin at roughly 4 times the pairing gap about 4 - 5 mev ) , we will underestimate the @xmath2b cross section .
phase space considerations reduce the importance of states this high in energy , however ; in this application , therefore , the error should be relatively small .
the final new element in this paper is the inclusion of spreading widths .
these depend on the density of complicated background states and on the average interaction strength coupling them to our model space , neither of which we can reliably estimate .
we settle instead for the prescription in ref.@xcite , developed primarily to describe spreading of rpa - like states high in energy , viz : @xmath45 \\ \gamma(\epsilon ) & = & 10.75 \left ( \frac{\epsilon^2}{\epsilon^2 + 18 ^ 2 } \right ) \left ( \frac{110 ^ 2}{\epsilon^2 + 110 ^ 2 } \right ) , \nonumber\end{aligned}\ ] ] where @xmath46 is the excitation energy , @xmath47 is a parameterization of the single - quasiparticle width ( in mev ) , and we have included a lower limit of 300 kev to simulate experimental resolution .
[ f : spread ] shows the full gt distribution with the widths included ( and divided by .76 _ in the figure only _ to account for the strength in five - quasiparticle states , so that the height of the giant resonance is roughly correct ) .
the spreading turns out to have a fairly small effect on both the total strength below threshold and the @xmath2b cross section .
our final values for these quantities , with all the new physics included and the spin operator quenched as discussed above , are 3.00 and 3.27@xmath3 .
the @xmath2b cross section is 50% larger than our old value .
using the free - nucleon axial - vector coupling constant would increase it another 60% and we can take that figure as a nominal measure of uncertainty in our result .
our strength distribution does not differ dramatically from the one extracted from experiment@xcite .
both suffer from uncertainty in the overall normalization , which , as discussed in the introduction , has prompted an attempt at lampf@xcite to measure the total cross section for absorbing electron neutrinos from muon - decay .
but the energies of these neutrinos extend to half the muon mass , much higher than the solar neutrino end point ( @xmath48 mev ) , and the momenta transferred are typically on the order of the inverse radius of the target nuclei . the allowed approximation , which ignores variations in the lepton wave functions over the interior of the nucleus , no longer applies and the cross section depends on nuclear wave functions in a more complicated way . ref .
@xcite contains a framework for treating neutrino - induced reactions beyond the allowed approximation ; it involves calculating matrix elements of the operators : @xmath49^j \nonumber \\
j_l(qr ) [ y_l(\omega_r ) \ , \frac{1}{m } \vec{\nabla } ] ^j \nonumber \\
j_j(qr ) y_j(\omega_r ) \ , \vec{\sigma } \cdot \frac{1}{m } \vec{\nabla}\end{aligned}\ ] ] where @xmath50 is the momentum transfer , @xmath51 is the @xmath52th spherical bessel function , @xmath53 is a spherical harmonic , @xmath54 is the multipole order of the transition , and @xmath55 is the nucleon mass .
the maximum value of @xmath56 is about 2.8 and we need not consider multipoles above @xmath57 . with our qtda wave functions the evaluation of the beam - dump cross section is nominally straightforward , but a few subtleties complicate the analysis .
the most important difficulty arises in connection with the coulomb attraction of the outgoing electron towards the @xmath58 xenon nucleus . in both
the low - energy ( @xmath59)@xcite and high - energy ( @xmath60)@xcite limits , clear procedures exist for evaluating coulomb effects ; in one case the outgoing electron is known to be in an @xmath61-state , enhancing the cross section by the usual fermi factor , and in the other the outgoing wave function can be treated as a plane wave perturbed slightly by the charge distribution in the nuclear interior . since @xmath62 here , neither limit applies and so we follow a different line of reasoning
. the multipolarity of the nuclear transition must be equal to that obtained from coupling the incoming and outgoing lepton angular momenta and parities .
as we shall show shortly , the most important transitions have multipole quantum numbers @xmath63 and @xmath64 . for @xmath63 transitions ,
the orbital angular momenta of the two leptons must be coupled to either 0 , or 2 .
the latter possibility requires both the neutrino and the electron to have @xmath65 ( or for one of them to have @xmath66 ) , which for the energies of interest here is less likely than the other possibility that they both have @xmath67 , i.e. that they are both in @xmath61-states .
we therefore treat coulomb effects in the @xmath63 channel as if the transitions were purely allowed that is , we multiply the @xmath63 cross section by the usual fermi function @xmath68 .
the @xmath64 transitions require at least one lepton to be in a @xmath69 ( or higher ) state , and the usual prescription makes less sense .
these are unique " transitions , however , and can be treated simply in processes such as beta decay@xcite , for which @xmath56 is always small .
the procedure there is to modify the usual cross section ( with fermi function ) by a factor @xmath70 where @xmath71 and @xmath72 are the neutrino and electron momenta , and @xmath73 is a function of @xmath72 tabulated in ref . @xcite .
though @xmath56 is not always small enough in our problem for the procedure to be strictly valid , we employ it to get a rough handle on the coulomb corrections ; they turn out to add very slightly to the corrections produced by the fermi function .
table [ t : multipole ] shows the contribution of each multipole to the beam - dump cross section , evaluated with the two commonly used values for @xmath1 ( -1.0 and -1.26 ) .
several effects emerge .
the presence of the bessel functions in the second , third , and fourth operators in eq .
[ e : ops ] , and of @xmath74 in the third and fourth operators , together reduce the @xmath63 contribution to about 2/3 of its allowed value .
( the bessel - function effects can be computed reliaby ; in fact , the same calculation in the simple helm model@xcite gives a very similar result . )
the contributions of the @xmath23 and @xmath75 operators , in which @xmath76 is the first bessel function to appear , are nominally suppressed by the upper limit in @xmath56 but are in fact far from negligible .
the @xmath77 , on the other hand , contributes little even though it is unsupressed because its strength is concentrated in the isobar analog state , which lies several mev above the neutron - emission threshold .
the @xmath78 and @xmath79 strengths are even more concentrated within the giant - dipole resonance and its spin analog , well above 7.23 mev .
in fact the @xmath64 contribution , which we find to be of the same order as that of the @xmath63 , may well be smaller than in our table , because the associated strength is probably also concentrated at higher energies .
our model space is clearly not large enough for the full collectivity of the spin - dipole mode to assert itself .
to get some idea of the size of our overestimate we performed qrpa calculations for @xmath80xe with the same single - particle levels and effective interaction we used in the odd - mass nuclei .
we then enlarged the model space to include the @xmath81 and @xmath82 levels a few mev above the fermi surface .
we found that while the total @xmath64 strength increased , the amount below 7 mev fell by 20 30% . including other levels below the fermi surface ( e.g. the @xmath83 shell )
would concentrate it further still ; just how much would remain below threshold is not clear .
altogether , the cross section for muon - decay neutrinos is only about half of the preliminary experimental value of 6@xmath84 cm@xmath42 @xcite , even without any quenching of the axial current .
there is no clear way to make our results compatible with this larger cross section .
as is apparent from the table , the bulk of the cross section comes from the @xmath63 and @xmath64 multipoles , and we have argued that the @xmath64 contribution is probably an overestimate . because the average neutrino energy is so much higher than 7 mev , phase space plays only a small role and the gamow - teller @xmath63 contribution depends essentially only on the total strength below the neutron - emission threshold it is insensitive to the precise distribution . to account for the difference between our calculated total cross section and a value of 6@xmath84 cm@xmath42
, the @xmath63 contribution would have to be about three times larger than in table [ t : multipole ] .
an increase of about 25% might be plausible since , as discussed in connection with fig .
[ f : spread ] , we have not included states with five or more quasiparticles . our claim , however , is that such states appear in large numbers only at energies above 4 5 mev and that relatively few will contribute to the strength below threshold . in our calculation
the summed gt strength below threshold is about 10% of the total strength , which is dominated ( @xmath86 70% ) by the states that form the giant gamow - teller resonance at 13 - 14 mev .
such a distribution is consistent with general experience in other nuclei ; in fact it seems impossible to increase the low - lying gt strength substantially as long as the giant gt resonance is at the energy suggested by the @xmath87 reaction , which in turn is in reasonable agreement with our calculation . finally ,
though it is not impossible that we are seriously underestimating the contributions of the other multipoles , that would not be good news either . a forbidden contribution even larger than our estimate would make an extraction of the gt distribution extremely difficult .
unfortunately , therefore , we can not reconcile our results with the value 6@xmath84cm@xmath42 without invoking an unforeseen mechanism that would probably spoil the calibration .
in summary , we have recalculated cross sections for the capture by @xmath0i of neutrinos from solar @xmath2b and @xmath4be decay incorporating several new physical effects and have found somewhat larger values than in our previous work .
the physics we still have not included , e.g. further correlations / deformation , a precisely determined and justified value for @xmath1 , a more carefully crafted interaction , etc .
, make our results somewhat uncertain , but with enough effort more comprehensive calculations are possible . our current results , we feel , still support the development of a calibrated iodine - based solar neutrino detector if only because it would be an improved version of the existing homestake experiment , which as of now is the most difficult to reconcile with standard - model physics .
we must also conclude , unfortunately , that a complete calibration may be more difficult than originally hoped .
we do not see how the preliminary lampf cross section can be due entirely to gamow - teller - like strength , and do not as yet know how to evaluate and remove the forbidden contributions with the required accuracy . nonetheless , we are hopeful that with the steady advance of experimental techniques , nuclear - structure expertise , and computing power , the remaining problems can be overcome .
we wish to acknowledge useful conversations with w.c .
haxton and k. lande .
this work was supported in part by the u.s .
department of energy under grants de - fg05 - 94er40827 and de - fg03 - 88er40397 , and by the national science foundation under grants phy-9108011 and phy-9303041 .
haxton , phys .
lett . * 60 * , 768 ( 1988 ) .
j. bahcall and h. bethe , phys .
rev . * d47 * , 1298 ( 1993 ) .
j. engel , s. pittel , and p. vogel , phys .
lett . * 67 * , 426 ( 1991 ) .
e.sugarbaker , private communication .
cleveland , t. daily , j. distel , k. lande , c.k .
lee et al . ,
proceedings of the 23rd international cosmic ray conference ( university of calgary , alberta , canada , 1993 ) vol .
3 , p. 865 .
a. hosaka , k.i .
kubo , and k. toki , nucl . phys . *
a444 * , 76 ( 1985 ) .
a. abzouzi , e. caurier , and a.p .
zucker , phys .
rev . lett . * 66 * , 1134 ( 1991 ) .
i. talmi , nucl . phys . *
a172 * , 1 ( 1971 ) .
k. grotz , h.v .
klapdor , and j. metzinger , phys .
* c * 33 , 1263 ( 1986 ) .
r.d . smith and j. wambach , phys .
* c38 * , 100 ( 1988 ) .
walecka , in _ muon physics _ , edited by v.w .
hughes and c.s .
wu ( academic , new york , 1975 ) , vol .
2 , p. 113 .
h. behrens and j. jnecke , in _ numerical tables for beta decay and electron capture , landolt - bornstein , new series group 1 _ , edited by k.h .
hellwege and h. schopper ( springer , new york , 1969 ) , vol .
4 . r. rosenfelder , ann
128 * , 188 ( 1980 ) .
h. schopper , _ weak interactions and nuclear beta decay _ , ( north - holland , amsterdam , 1966 ) . h. behrens and w. bhring , _ electron radial wave functions and nuclear beta - decay _ ( clarendon , oxford , 1982 ) . r.h .
helm , phys .
rev . * 104 * , 1466 ( 1956 ) .
.energies of low - lying states in @xmath0i and @xmath0xe in kev .
the second column contains measured values , the third the qtda predictions , and the last the gs predictions .
neither calculation incorporates three - like - quasiparticle states ( or their analogs in the gs scheme ) .
[ t : energies ] [ cols="^,^ , > , > , > " , ] | we discuss and improve a recent treatment of the absorption of solar neutrinos by @xmath0i , in connection with a proposed solar neutrino detector . with standard - solar - model fluxes and an in - medium value of -1.0 for the axial - vector coupling constant @xmath1
, we obtain a @xmath2b - neutrino cross section of 3.3@xmath3 , about 50% larger than in our previous work , and a @xmath4be cross section that is less certain but nevertheless also larger than before .
we then apply the improved techniques to higher incoming energies that obtain at the lampf beam dump , where an experiment is underway to finalize a calibration of the @xmath0i with electron neutrinos from muon decay .
we find that forbidden operators , which play no role in solar - neutrino absorption , contribute nonnegligibly to the lampf cross section , and that the preliminary lampf mean value is significantly larger than our prediction . |
the sieve of eratosthenes is the oldest and most well - known of the integer sieves , and is used to find all the primes up to a given limit @xmath5 .
the sieve begins with the list of integers @xmath6 and proceeds iteratively by marking the smallest number on the list as prime and removing it along with its multiples from the list .
the smallest number still left on the list is marked as prime and the procedure continues until the list is empty .
algorithmically , the sieve of eratosthenes both identifies the prime numbers in the list and yields a unique prime factorization for the composite numbers through multiple presentations of each polynomial value as product of two integers . in other words ,
each value @xmath7 in the sequence @xmath8 is presented as the factorization presentation @xmath9 for each @xmath10 .
if however @xmath11 is an arbitrary polynomial with integer coefficients and @xmath10 , then @xmath12 for each @xmath13 too .
hence , the algorithm can be generalized to include other polynomials at the cost of missing some of the factorization presentations .
fortunately , the situation can be improved by taking both factors of each composite @xmath14 into consideration , i.e. , if @xmath15 is marked as being divisible by @xmath16 then all @xmath17 where @xmath13 can be marked as being divisible by @xmath18 as well . to keep track of all the factorization presentations
, it suffices to record the initial value along with the sequence of quotients for the multiples of the factors , e.g. , if @xmath19 , @xmath20 and @xmath21 then the factorization presentation can be reconstructed from the sequence @xmath22 .
this method of sieving the polynomial values for integer factorizations is expressed in theorem [ main_recurrence ] , and holds in the context of multivariate polynomials as well .
section [ sec : recursively - factorable ] introduces a family of polynomials called _ recursively - factorable polynomials _ for which the collection of factorization presentations corresponding to the sequences @xmath23 yield the unique prime factorization for each value of @xmath11 via presentations @xmath9 for each @xmath10 . in general ,
recursively - factorable polynomials are rare , but there are some noteworthy instances .
particularly , the _ euler - like _ and _ legendre - like _ prime producing polynomials of the form @xmath24 for @xmath25 and @xmath26 for @xmath27 , respectively , and landau s @xmath0 are recursively - factorable .
the sieve of eratosthenes verifies that the line @xmath28 is also recursively - factorable , but we presently focus on recursively - factorable quadratic equations . in section [ sec : quad_dio ]
, we introduce an identity which presents the factorization of a quadratic polynomial value as the product of two binary quadratic forms ( theorem [ thm : dio_main ] ) and show that this identity associates all the factorization presentations of the aforementioned polynomial - value sieving integer sequences with the set @xmath29 . for monic quadratics , @xmath30 and the factorization presentations correspond to the transvection generators of @xmath31 ( corollary [ cor : transvections ] ) . in section [ sec : lattice_conic ] , a bijection is established ( theorem [ thm : conic_bijection ] ) between @xmath32 and the set @xmath33 of lattice point solutions @xmath34 for the conic sections @xmath35 with @xmath36 , showing that @xmath33 does not depend on @xmath37 , @xmath38 , or @xmath28 . following the mappings in figure [ fig : outline ] , each lattice point @xmath39 of the conic section is associated with an element of @xmath32 and gives a factorization presentation for @xmath40 .
if a factorization presentation @xmath9 has a corresponding integer sequence @xmath41 then there is a matching element of @xmath32 which corresponds to a lattice point solution of the conic section .
( magic ) [ matrix of nodes , ampersand replacement= & , column sep=3.5em ] @xmath9 & & + & @xmath42 & @xmath43 + @xmath41 & & + ; ( magic-2 - 2 ) edge [ bend left=15 ] node[auto ] thm .
[ thm : conic_bijection ] ( magic-2 - 3 ) ( magic-2 - 3 ) edge [ bend left=15 ] node[auto ] thm .
[ thm : conic_bijection ] ( magic-2 - 2 ) ( magic-2 - 2 ) edge [ bend right=15 ] node[above ] thm .
[ thm : dio_main ] ( magic-1 - 1 ) ( magic-3 - 1 ) edge [ bend right=15 ] node[below ] thm .
[ thm : dio_quad ] ( magic-2 - 2 ) ( magic-3 - 1 ) edge [ bend left=15 ] node[auto ] thm .
[ main_recurrence ] ( magic-1 - 1 ) ; ( magic-1 - 1 ) edge [ bend left=15 ] node[auto ] thm .
[ thm : all_factors ] ( magic-3 - 1 ) ;
[ main_recurrence ] let @xmath44 be a commutative ring with identity . for any polynomial @xmath45 $ ] of degree @xmath46
, there exists a sequence of multivariate polynomials @xmath47 such that @xmath48 $ ] and @xmath49 where @xmath50 , @xmath51 , and @xmath52 for @xmath53 with the convention that @xmath54 is shorthand for @xmath55 .
since @xmath56 represents the trivial factorization , the statement is initially true and we proceed by induction on @xmath57 .
let @xmath58 be the @xmath59th order hasse derivative and @xmath60 be the @xmath59th order hasse derivative with respect to the intermediate @xmath61 .
applying @xmath62 to both sides of @xmath63 gives @xmath64 using the taylor series expansion for @xmath11 , @xmath65 which gives a definition for @xmath66 $ ] . substituting ( [ partial_f ] ) into ( [ hasse_taylor ] ) yields
@xmath67 [ quad_recurrence ] for @xmath68 , taking @xmath69 in equation ( [ partial_f ] ) gives @xmath70 for all @xmath71 .
so for @xmath72 , theorem [ main_recurrence ] expresses @xmath54 as @xmath73 for each sequence @xmath74 , if @xmath75 then @xmath76 moreover if @xmath77 , then there exists @xmath78 such that @xmath79 where @xmath80 and @xmath81 .
[ recur_ex ] let @xmath82 .
we compute @xmath83 as follows : @xmath84 this gives @xmath85 .
one can also verify that @xmath86
theorem [ main_recurrence ] provides a means of factoring the values of a polynomial @xmath11 into two integers , but these presentations may not represent the full solution set @xmath87 .
for example when @xmath88 , the integer factorization @xmath89 can not be presented via theorem [ main_recurrence ] , i.e. , there does not exist a finite sequence of integers @xmath90 for which @xmath91 , @xmath92 , and @xmath93 .
proof of this fact is shown in remark [ non - recur - fac ] . by contrast , lemma [ lem : recur_exist ] provides the existence of a family of polynomials @xmath94 for which the prime integer factorization of each value of @xmath95 can be reconstructed from the presentations of theorem [ main_recurrence ] .
theorem [ thm : all_factors ] shows that this family of polynomials contains the recursively - factorable polynomials characterized by the following property .
let @xmath11 be a polynomial with integer coefficients .
if for each integer factorization presentation @xmath9 there exists an @xmath96 such that @xmath97 and @xmath98 or @xmath99 , then @xmath28 is said to satisfy the _
recursively - factorable criterion _ for @xmath11 .
if each @xmath100 satisfies the recursively - factorable criterion for @xmath11 , then the polynomial @xmath11 is said to be _ recursively - factorable_. recursively - factorable polynomials are irreducible over @xmath101 . if not then @xmath102 for some @xmath100 , but the non - trivial factorization @xmath103 has no associated @xmath104 such that @xmath105 for any @xmath106 .
[ lem : shift_still_recursive ] let @xmath11 be a polynomial and @xmath107 for some @xmath108 .
if @xmath11 is recursively - factorable , then so is @xmath109 .
suppose that @xmath110 is a non - trivial factorization .
since @xmath11 is recursively - factorable , we may assume without loss of generality that there exists @xmath111 such that @xmath112 where @xmath113 .
thus @xmath114 and @xmath115 , so we may conclude that @xmath109 is recursively - factorable .
[ thm : all_factors ] if @xmath11 is recursively - factorable then , for each @xmath116 and @xmath117 such that @xmath10 , there exists a finite sequence of integers @xmath90 such that @xmath118 fix @xmath100 . if @xmath119 then the sequence @xmath120 gives the presentation @xmath121 .
thus it is sufficient to consider the case where @xmath14 is a composite integer with a non - trivial factorization @xmath122 such that @xmath123 .
let @xmath124 .
since @xmath11 is recursively - factorable , there exists an @xmath125 such that @xmath97 .
moreover there is an @xmath126 such that @xmath127 for all @xmath125 . set @xmath128 so that @xmath129 .
it follows that @xmath130 and @xmath131 for some @xmath132 . if @xmath133 , then @xmath131 represents a trivial factorization and the sequence @xmath134 yields the presentation @xmath135 if @xmath136 , then @xmath137 represents a non - trivial factorization , and by the minimality of our choice of @xmath138 relative to all other @xmath125 there exists an @xmath139 which minimizes @xmath140 over all @xmath141 , i.e. , @xmath142 for some @xmath143
. we may continue in this fashion until we obtain the trivial integer factorization @xmath144 where @xmath145 , produced from a finite sequence of factors @xmath146 , quotients @xmath147 and remainders @xmath148 such that @xmath149 and @xmath150 for each @xmath151 . starting with @xmath152 and @xmath144
we may reverse this sequence to obtain @xmath28 and @xmath16 as follows : @xmath153 p_{m-2 } & = \frac{f(r_{m-2})}{p_{m-1 } } = \frac{f(r_{m-1 } + q_{m-1 } \
, p_{m-1})}{p_{m-1 } } = \frac{f(r_{m-1 } \ , f_0 + q_{m-1 } \ ,
f_1 ( r_{m-1}))}{f_1 ( r_{m-1 } ) } = f_2 ( r_{m-1 } , q_{m-1}).\end{aligned}\ ] ] more generally @xmath154 for @xmath155 and @xmath156 .
therefore the integer sequence @xmath157 gives the presentation @xmath158 and @xmath159 .
smaller_factors.pdf ( 99,67)@xmath160 ( -8,48)@xmath161 ( 29,37.5)@xmath162 ( 68,22)@xmath163 ( 55,13)@xmath164 ( 18,10)@xmath165 ( 70,4)@xmath166 the proof of theorem [ thm : all_factors ] starts with an integer factorization @xmath167 and constructs a sequence of factorizations @xmath168 , @xmath169 such that @xmath170 until a prime number @xmath171 with the trivial factorization @xmath172 is reached . in this way
prime - producing polynomials , which contain a large interval of consecutive prime values , make good candidates for having the recursively - factorable property . in 1772 , euler @xcite discovered that the polynomial @xmath173 produces prime numbers for @xmath174 $ ] , and later legendre @xcite noted that both @xmath175 and @xmath1 are prime for @xmath176 $ ] and @xmath177 $ ] , respectively . le lionnais considered polynomials of the type @xmath178 in general , which he called euler - like polynomials @xcite , and integers @xmath179 for which @xmath178 is prime for @xmath180 have come to be known as _
lucky numbers of euler_. rabinowitz @xcite proved that @xmath179 is a lucky number of euler if and only if the field @xmath181 has class number 1 . from this
, heegner @xcite and stark @xcite showed that there are exactly six lucky numbers of euler , namely 2 , 3 , 5 , 11 , 17 , and 41 .
legendre @xcite explored other types of prime - producing quadratics such as @xmath182 which is prime when @xmath183 for @xmath184 .
akin to the euler - like polynomials , these quadratics give primes for @xmath185 for prime @xmath186 if and only if @xmath187 has class number 2 @xcite .
baker @xcite and stark @xcite found that the only such @xmath186 are 3 , 5 , 11 , and 29 .
as seen in lemma [
lem : recur_exist ] , euler - like and legendre - like prime - producing quadratics are indeed recursively - factorable . further discussion of prime - producing quadratics can be found in @xcite . [
lem : recur_exist ] the following quadratics ( and their horizontal shifts ) are recursively - factorable :
@xmath188 where @xmath189 , 2 .
@xmath190 where @xmath191 [ euler - like ] 3 .
@xmath26 where @xmath192 , [ legendre - like ] 4 .
@xmath193 where @xmath194 , [ 2n^2 + 2n - like ] 5 .
@xmath195 where @xmath196 , 6 .
@xmath197 where @xmath198 , 7 .
@xmath199 where @xmath200 , and 8 .
@xmath201 where @xmath202 .
we claim that if @xmath11 is one of these polynomials and all the values within a suitably large interval @xmath203 are known to satisfy the recursively - factorable criterion for @xmath11 , then the remaining values outside of @xmath203 also satisfy the recursively - factorable criterion . supposing @xmath40 is one of the polynomials in cases ( i)-(viii ) , @xmath11 is a positive parabola having a minimum at either @xmath204 or @xmath205 .
furthermore the values @xmath206 for all @xmath100 , so if @xmath28 satisfies the recursively - factorable criterion then so does @xmath207 .
also note that @xmath208 for @xmath209 . for cases ( i)-(vi ) , define @xmath210 such that @xmath211 for each @xmath212 .
given that for each factorization presentation @xmath213 either @xmath214 or @xmath215 , for @xmath212 there exists a @xmath13 such that either @xmath216 or @xmath217 .
thus if we can verify that the values within @xmath203 satisfy the recursively - factorable criterion , then so do the values greater than @xmath210 ( and symmetrically the values less than @xmath218 ) , i.e. , @xmath11 is recursively - factorable . in cases ( vii ) and ( viii ) we use a sharper approximation of @xmath219 than @xmath220 to determine @xmath210 , but the idea is the same . in cases ( i ) , ( iii ) , ( v ) , and ( vii ) , @xmath14 is prime ( or 1 ) for @xmath221 $ ] and @xmath222 which means @xmath223 , so the recursively - factorable condition is satisfied for @xmath224 $ ] .
similarly , @xmath14 is prime ( or 1 ) for @xmath225 $ ] in cases ( ii ) , ( iv ) , ( vi ) , and ( viii ) .
the recursively - factorable condition is satisfied for @xmath226 , @xmath227 , and @xmath38 since @xmath228 and @xmath229 .
hence for all cases ( i)-(viii ) the recursively - factorable criterion is satisfied for @xmath224 $ ] . : for @xmath230 with @xmath189 , @xmath231 and @xmath232 \subset [ -c , c]$ ] . : for @xmath190 with @xmath233 , @xmath234 $ ] and yields the respective @xmath203 intervals corresponding to each @xmath38 : @xmath235 \subseteq [ -1,1]$ ] , @xmath236\subseteq [ -2,2]$ ] , @xmath236\subseteq [ -3,3]$ ] , @xmath236\subseteq [ -5,5]$ ] , @xmath237\subseteq [ -11,11]$ ] , @xmath237\subseteq [ -17,17]$ ] , and @xmath238\subseteq [ -41,41]$ ] . :
for @xmath239 with @xmath240 , @xmath241 $ ] which gives the respective intervals : @xmath242 \subseteq [ -1,1]$ ] , @xmath243 \subseteq [ -3,3]$ ] , @xmath243 \subseteq [ -5,5]$ ] , @xmath244 \subseteq [ -11,11]$ ] , and @xmath245 \subseteq [ -29,29]$ ] .
: let @xmath246 with @xmath194 , @xmath247 $ ] which gives the respective intervals : @xmath242 \subseteq [ -1 , 1]$ ] , @xmath242 \subseteq [ -2 , 2]$ ] , @xmath242 \subseteq [ -3 , 3]$ ] , @xmath243 \subseteq [ -7 , 7]$ ] , and @xmath248 \subseteq [ -19 , 19]$ ] .
: let @xmath249 , then @xmath250 \subseteq [ -2,2]$ ] .
: let @xmath251 with @xmath198 , @xmath252 $ ] which gives the respective intervals : @xmath235 \subseteq [ -1 , 1]$ ] , @xmath235 \subseteq [ -2 , 2]$ ] , @xmath237 \subseteq [ -5 , 5]$ ] , @xmath253 \subseteq [ -11 , 11]$ ] , and @xmath254 \subseteq [ -23 , 23]$ ] .
: let @xmath11 be of the form @xmath199 with @xmath200 .
we claim that if @xmath9 where @xmath255 is an integer factorization presentation , then @xmath256 .
observe that @xmath257 implies that @xmath258 and @xmath259 and is a contradiction for @xmath260 .
similarly , for @xmath261 and @xmath262 , @xmath263 and is also contradiction for @xmath260 .
clearly @xmath264 since @xmath265 .
thus we are guaranteed that @xmath266 and there exists an @xmath267 such that @xmath268 .
: let @xmath11 be of the form @xmath201 with @xmath202 .
as in case ( vii ) , we show that @xmath269 for each integer factorization presentation @xmath213 where @xmath255 .
first notice that taking @xmath270 and @xmath271 leads to @xmath272 and is a contradiction for @xmath260 .
likewise , taking @xmath257 and @xmath271 gives @xmath273 and again is a contradiction for @xmath260 . with @xmath274 and @xmath275 , @xmath276 as @xmath277 .
finally assume that @xmath261 and @xmath278 , @xmath279 and is a contradiction for @xmath260 .
finally take @xmath280 to get the contradiction @xmath281 .
therefore if the recursively factorable criterion holds for the values in the interval @xmath282 $ ] , then @xmath266 and the criterion holds for the values outside of the interval also .
.recursively - factorable polynomials with real roots . [ cols="<,<",options="header " , ] with some additional casework to show that the values over a suitably large interval satisfy the recursively - factorable criterion , it can also be shown that the polynomials in table [ tab : real_roots_recur_fac ] are recursively - factorable
. some of these quadratics are prime - producing polynomials , or a horizontal shift of one , listed in @xcite and @xcite . for these real - root quadratics , the condition @xmath283 for @xmath284 $ ] no longer holds as it did in lemma [ lem : recur_exist ] .
however for @xmath285 , @xmath208 for all @xmath286 . hence @xmath210 can be chosen to be sufficiently large so that , for all @xmath287 , both @xmath208 for @xmath288 and @xmath289 .
we show in this section that , for quadratic polynomials , the factorization presentations of theorem [ main_recurrence ] , defined recursively as @xmath290 , may be expressed in a closed form as the product of two binary quadratic forms .
theorem [ thm : dio_quad ] establishes that , in this context , each factorization presentation sequence @xmath291 corresponds with a particular @xmath292 .
fix @xmath293 .
let @xmath294 , @xmath295 , @xmath296 , and @xmath297 be functions from @xmath298 defined such that for @xmath299 , @xmath300 & = \alpha \ , \delta - a \ , \beta \ , \gamma , \\
\eta_f [ a ] & = \alpha \ , \gamma + b \ , \beta \ ,
\gamma + c \ , \beta \ , \delta , \\ \phi_{f,0 } [ a ] & = { \alpha}^2 + b \ , \alpha \ , \beta + a \ , c \ , { \beta}^2,\\ \phi_{f,1 } [ a ] & = a { \gamma}^2 + b \ ,
\gamma \ , \delta + c \ , { \delta}^2 , \end{aligned}\ ] ] and for natural @xmath57 , @xmath301 = \begin{cases}\phi_{f,0}[a ] & \mbox{for even } m \\ \phi_{f,1}[a ] & \mbox{for odd } m \end{cases } .\ ] ] we suppress the @xmath11 when it is clear by the context , favoring the notation @xmath302 $ ] , @xmath303 $ ] , @xmath304 $ ] , @xmath305 $ ] , and @xmath306 $ ] . for @xmath307 ,
let @xmath308 = 1 \right\ } .\ ] ] in general , the set @xmath32 is not closed under matrix multiplication and does not contain its inverses .
however the case when @xmath30 is particularly noteworthy as @xmath31 is the special linear group .
[ thm : dio_main ]
let @xmath309 such that @xmath310 .
for @xmath311 , @xmath312 if and only if @xmath313 or @xmath314 , i.e. , for @xmath315 , @xmath316 ) = \phi_0 [ a ] \ , \phi_1 [ a]\ ] ] if and only if @xmath317 = 1 $ ] or @xmath318 $ ] .
by expanding both sides , one can verify that : @xmath319 the set of matrices @xmath320 given by @xmath321 @xmath322 and @xmath323 given by @xmath324 respectively , correspond to the trivial factorization in theorem [ thm : dio_main ] for each @xmath325 .
the fibonacci - brahmagupta identity has a long history in mathematics beginning with its first appearance in diophantus _ arithmetica _ ( iii , 19 ) @xcite c.250 in the form of @xmath326 . later in c.628 , brahmagupta generalized diophantus identity by showing that numbers of the form @xmath327 are closed under multiplication .
brahmagupta s identity was popularized in 1225 upon its reprinting in fibonacci s _ liber quadratorum _
@xcite where the first rigorous proof of the identity appeared . finally in 1770 , euler @xcite further generalized brahmagupta s identity by providing the parametric solution @xmath328 for the diophantine equation @xmath329 with composite @xmath330 . in corollary [ fib - brahm_identity ] we show that the case @xmath331 in theorem [ thm : dio_main ] corresponds to the case @xmath332 in euler s identity ( [ euler_identity ] ) .
[ fib - brahm_identity ] @xmath333 when @xmath331 , @xmath334 and @xmath335 where @xmath313 .
hence @xmath336 [ thm : dio_quad ] for @xmath293 and @xmath337 , @xmath338\ ] ] where @xmath339 defined recursively by @xmath340 for @xmath341 such that @xmath342 we shall proceed by induction on @xmath57 . for each @xmath343 , define @xmath344 and @xmath345 recursively as stated in the hypothesis . initially we see that @xmath346 $ ] and @xmath347 $ ] satisfies the hypothesis .
now assume @xmath348 $ ] and @xmath349 $ ] for each @xmath350 .
suppose @xmath351 for some @xmath352 .
remark [ quad_recurrence ] gives @xmath353 + a \ , x_{2j}^2 f_{2j-1}.\ ] ] by the induction hypothesis @xmath354 = \alpha_{2j-2}^2 + b \
, \alpha_{2j-2 } \
, \beta_{2j-2 } + a c \ , \beta_{2j-2}^2\ ] ] and @xmath355 = a \ , \gamma_{2j-1}^2 + b \ , \gamma_{2j-1 } \
, \delta_{2j-1 } + c \ , \delta_{2j-1}^2.\ ] ] the partial derivative @xmath356 \big]$ ] may be evaluated through the equation @xmath357 .
in particular @xmath358 = \alpha_{2j-2 } \quad \mbox{and } \quad { \frac{\partial^{\!}}{\partial { x_{2j-1}}^{\ ! } } } \big[\beta_{2j-1}\big ] = a \ , \beta_{2j-2}\ ] ] which yields @xmath359 & = { \frac{\partial^{\!}}{\partial { x_{2j-1}}^{\ ! } } } \big[\phi_1 [ a_{2j-1 } ] \big ] \\ & = { \frac{\partial^{\!}}{\partial { x_{2j-1}}^{\ ! } } } \left [ a \ , \gamma_{2j-1}^2 + b \ , \gamma_{2j-1 } \delta_{2j-1 } + c \ , \delta_{2j-1}^2 \right ] \\ & = 2 \ , a \ , \gamma_{2j-1 } \ , \alpha_{2j-2 } + b \left ( a \ , \gamma_{2j-1 } \ , \beta_{2j-2 } + \delta_{2j-1 } \ , \alpha_{2j-2 }
\right ) + 2 \ , a c \ , \delta_{2j-1 } \ , \beta_{2j-2 } \end{aligned}\ ] ] substituting ( [ quad_fac_2j-2 ] ) , ( [ quad_fac_2j-1 ] ) , and ( [ quad_der_2j-1 ] ) into ( [ quad_partial_2j ] ) gives @xmath360 as defined in the hypothesis , @xmath361 so @xmath362 & = \phi_0 [ a_{2j}]\\ & = ( \alpha_{2j-2 } + a \ , x_{2j } \ , \gamma_{2j-1})^2 + ac \ , ( \beta_{2j-2 } + x_{2j } \ , \delta_{2j-1})^2 \\ & \quad + b \ , ( \alpha_{2j-2 } + a \ , x_{2j } \ , \gamma_{2j-1})(\beta_{2j-2 } + x_{2j } \ , \delta_{2j-1 } ) .
\end{aligned}\ ] ] comparing ( [ fac_comp_2j ] ) and ( [ phi_comp_2j ] ) shows that @xmath363 $ ] .
initially @xmath364 = 1 $ ] and by the induction hypothesis @xmath365 = 1 $ ] for @xmath366 , so we check that @xmath339 : @xmath367 & = \delta \left[\begin{pmatrix}\alpha_{m-1 } + x_m \ , a \ , \gamma_{m-1 } & \beta_{m-1 } + x_m \ , \delta_{m-1 } \\
\gamma_{m-1 } & \delta_{m-1 } \end{pmatrix } \right ] \\ & = ( \alpha_{m-1}+x_m \ , a \ , \gamma_{m-1 } ) \ , \delta_{m-1 } - a \ , ( \beta_{m-1 } + x_m \ , \delta_{m-1 } ) \
, \gamma_{m-1 } \\ & = ( \alpha_{m-1 } \ ,
\delta_{m-1 } - a \ , \beta_{m-1 } \ , \gamma_{m-1 } ) = \delta[a_{m-1 } ] = 1 \ , .\end{aligned}\ ] ] similarly when @xmath368 , remark [ quad_recurrence ] says that @xmath369 + a \ , x_{2j+1}^2 f_{2j}.\ ] ] whose partial derivative @xmath370 = { \frac{\partial^{\!}}{\partial { x_{2j}}^{\ ! } } } \big [ \phi_{2j } [ a_{2j } ] \big]$ ] may be computed through ( [ phi_comp_2j ] ) as @xmath371 & = 2 \ , a \ ,
\alpha_{2j } \
, \gamma_{2j-1 } + b \ , \alpha_{2j } \ , \delta_{2j-1 } + a \ , b \ , \gamma_{2j-1 } \ , \beta_{2j } + 2
\ , a \ , c \ , \beta_{2j } \ , \delta_{2j-1 } \end{aligned}\ ] ] since @xmath372 and @xmath373 . putting ( [ quad_fac_2j-1 ] ) , ( [ quad_partial_2j+1 ] ) , and ( [ quad_der_2j ] ) together with the fact that @xmath374 $ ] gives @xmath375 and may be compared with @xmath376 $ ] which is computed thusly : @xmath377 & = \phi_1 \left [ \begin{pmatrix } \alpha_{2j-1 } + a \ , x_{2j } \
, \gamma_{2j-1 } & \beta_{2j-1 } + x_{2j } \ , \delta_{2j-1 } \\
\gamma_{2j-1 } + x_{2j+1 } \ , \alpha_{2j } & \delta_{2j-1 } + a \ , x_{2j+1 } \ , \beta_{2j } \end{pmatrix } \right ] \\ & = a \ , ( \gamma_{2j-1 } + x_{2j+1 } \ , \alpha_{2j})^2 + c \ , ( \delta_{2j-1 } + a \ , x_{2j+1 } \ , \beta_{2j})^2 \\ & \quad + b\ , ( \gamma_{2j-1 } + x_{2j+1 } \ , \alpha_{2j})(\delta_{2j-1 } + a \ , x_{2j+1 } \ ,
\beta_{2k } ) .
\end{aligned}\ ] ] checking that ( [ fac_comp_2j+1 ] ) is equal to ( [ phi_comp_2j+1 ] ) shows @xmath378 $ ] .
we have that @xmath379 = 1 $ ] for @xmath341 , so @xmath367 & = \delta \left[\begin{pmatrix}\alpha_{m-1 } & \beta_{m-1 } \\ \gamma_{m-1 }
+ x_m \ , \alpha_{m-1 } & \delta_{m-1 } + x_m \ , a \ , \beta_{m-1 } \end{pmatrix } \right ] \\ & = \alpha_{m-1 } \ , ( \delta_{m-1}+x_m \ , a \ , \beta_{m-1 } ) - a \ , \beta_{m-1 } ( \gamma_{m-1}+ x_m \ , \alpha_{m-1 } ) \\ & = ( \alpha_{m-1 } \ , \delta_{m-1 } - a \ , \beta_{m-1 } \ , \gamma_{m-1 } ) = \delta[a_{m-1 } ] = 1 \ , .\end{aligned}\ ] ] which completes the proof . combining theorems [ thm : all_factors ] and [ thm : dio_quad ] implies that for a recursively - factorable polynomial @xmath11 , each non - trivial factorization presentation @xmath380 is represented by some @xmath339 via the identity @xmath381)=\phi_0 [ a_m ] \ , \phi_1 [ a_m]$ ] from theorem [ thm : dio_main ] . returning to example [ recur_ex ] , for @xmath382 we can compute @xmath83 using theorem [ thm : dio_quad ] : @xmath383 and @xmath384 = 3 \ , ( -18)^2 + 5 \
, ( -18 ) ( -11 ) + 11 \ , ( -11)^2 = 3293.\ ] ] it is readily checked that @xmath385 = 1 $ ] and meets the conditions of theorem [ thm : dio_main ] . since @xmath386 = 301 $ ] and @xmath387 = 83 $ ] , it follows that @xmath388 [ non - recur - fac ] the non - trivial factorization @xmath89 , but @xmath389 is the only value less than @xmath390 and @xmath391 .
likewise @xmath392 can not be represented by theorem [ thm : dio_main ] , since @xmath393 can not be represented by the binary form @xmath394 = \alpha^2 + \alpha \ , \beta + 7 \
, \beta^2 $ ] , see @xcite for more details .
recall that the special linear group may be generated by its transvections @xcite .
in particular , @xmath395 where @xmath396 and @xmath397 .
it follows that @xmath398 for all @xmath399 .
[ cor : transvections ] for @xmath400 , @xmath401\ ] ] where @xmath402 from theorem [ thm : dio_quad ] , @xmath378 $ ] where @xmath403 and @xmath404 for each @xmath343 .
it stands to reason that shifting a polynomial horizontally does not change the integer factorization of its values . in the case of quadratics ,
the specific correspondence between a parabola and its shift is expressed by the following proposition .
parabola_shifted.pdf ( 73,55)@xmath11 ( 93.5,55)@xmath109 ( 73,42)@xmath405 ( 39,39)@xmath406 \ , \phi_{f,1}[a]$ ] ( 88,39)@xmath407 \ , \phi_{g,1}[b]$ ] ( 58,1)@xmath408 $ ] ( 78,1)@xmath409 $ ] [ horz_shift ] let @xmath293 and set @xmath410 for some @xmath108 . for each @xmath411
there is a corresponding @xmath412 for which the following conditions hold : 1 .
@xmath413 , 2 .
@xmath414 = \eta_f [ a ] + h$ ] , 3 . @xmath415 = \phi_{f,0 } [ a]$ ] , and 4 .
@xmath416 = \phi_{f,1 } [ a]$ ] .
let @xmath417 such that @xmath418 .
noting that @xmath419 @xmath420 & = ( \alpha + h \ , a \beta ) \
, \delta - a \ , \beta ( \gamma + h \ , \delta ) \tag{i } \\ & = \alpha \delta - a \ , \beta \gamma = 1 .
\\[6pt ] \eta_g [ b ] & = ( \alpha + h \ , a \beta ) ( \gamma + h \ , \delta ) + ( b - 2 a h ) \ , \beta ( \gamma + h \ , \delta ) + ( c - b h + a h^2 ) \ , \beta \delta \tag{ii } \\
& = ( \alpha \gamma + b \ , \beta \gamma + c \ , \beta \delta ) + h \ , ( \alpha \delta - a \beta \gamma ) \\ & = \eta_f [ a ] + h .
\\[6pt ] \phi_{g,1 } [ b ] & = a \ , ( \gamma + h \ , \delta)^2 + ( b - 2 a h ) ( \gamma + h \ , \delta ) \delta + ( c - b h + a h^2 ) \ , \delta^2 \tag{iii } \\ & = a \ , \gamma^2 + b \ , \gamma \delta + c \ , \delta^2 .
\\[6pt ] \phi_{g,2 } [ b ] & = ( \alpha + h \ , a \beta)^2 + ( b - 2 a h ) ( \alpha + h \ , a \beta ) \beta + a ( c - b h + a h^2 ) \ , \beta^2 \tag{iv } \\ & = \alpha^2 + b \ , \alpha \beta + a c \ , \beta^2 .
\qedhere\end{aligned}\ ] ]
lastly , theorem [ thm : conic_bijection ] relates the set @xmath32 with the lattice point solutions of the conic sections @xmath422 . from theorem [ thm :
dio_main ] , each @xmath339 corresponds to an integer factorization presentation of a value of @xmath423 , i.e. , the problem of finding lattice point solutions to these conic sections is equivalent to factoring the value of an associated quadratic polynomial .
[ thm : conic_bijection ] for @xmath424 , let @xmath425 the map @xmath426 defined by @xmath427 is a bijection .
fix @xmath424 and consider @xmath428 . set @xmath429
$ ] , @xmath430 , @xmath431 , and @xmath432 .
direct substitution shows that @xmath433 = n. \label{ellipse_1}\ ] ] since @xmath434 , it follows that @xmath302=1 $ ] and @xmath435 , i.e. , @xmath436 solving for @xmath437 in ( [ ellipse_1 ] ) and substituting it into ( [ ellipse_2 ] ) shows that @xmath39 is a solution to @xmath438 now consider the inverse map @xmath439 defined by @xmath440 for each @xmath441 , @xmath442 = 1 $ ] and from ( [ ellipse_3 ] ) @xmath443 so @xmath444 .
hence @xmath445 .
we show that @xmath446 is injective by verifying that @xmath447 for each @xmath434 . indeed , since @xmath302=1 $ ] the @xmath448 implying that @xmath449 , i.e. , @xmath450 .
thus , @xmath451 = \psi^{-1 } \left [ \begin{pmatrix } \beta \gamma \\
\beta \delta \end{pmatrix } \right ] = \begin{pmatrix } \frac{\beta}{\beta \delta}(1 + a \beta \gamma ) & \beta \\
\frac{\beta \gamma}{\beta } & \frac{\beta \delta}{\beta } \end{pmatrix } = a\ ] ] since @xmath302=1 $ ] implies that @xmath452 . likewise , for each @xmath453 ,
@xmath454 = \psi \left [ \begin{pmatrix } \frac{g}{y}(1 + a x ) & g \\ \frac{x}{g } & \frac{y}{g } \\
\end{pmatrix } \right ] = \begin{pmatrix } x \\ y \end{pmatrix}\ ] ] meaning @xmath446 is surjective .
the mapping @xmath455 defined by @xmath456 = \begin{pmatrix } \beta \gamma \\ \beta \delta \end{pmatrix}$ ] is well - defined and onto , but is not one - to - one .
similarly , when @xmath30 or @xmath457 the respective mappings @xmath458 and @xmath459 are onto but not one - to - one .
therefore the image of @xmath446 under @xmath32 is @xmath33 . for @xmath460 .
the case @xmath461 is highlighted in blue and lattice points @xmath462 intersecting the ellipses are indicated.,width=480 ] consider the euler - like polynomial @xmath463 .
it is easy to verify that @xmath464 is a solution of @xmath465 by theorem [ thm : conic_bijection ] , the point @xmath466 corresponds to the element @xmath467 given by @xmath468 = \begin{pmatrix } 1 & 1\\ 3 & 4 \end{pmatrix}.\ ] ] thus @xmath469 ) = f(20 ) = 5 \cdot 77 = \phi_1 [ a ] \phi_2 [ a]$ ] .
similarly @xmath470 , @xmath471 , @xmath472 , @xmath473 , @xmath474 , @xmath475 and @xmath476 are also lattice point solutions ( see figure [ ellipse_plot ] ) to ( [ conic_example ] ) corresponding to the integer factorizations @xmath477 , @xmath478 , @xmath479 , @xmath480 , @xmath481 , @xmath482 , and @xmath483 , respectively .
gauss @xcite showed that the general binary quadratic diophantine equation can be reduced to a special case of the pell equation .
in particular , ( [ ellipse_3 ] ) can be reduced to @xmath484 where @xmath485 and @xmath486 provided that @xmath487 .
the trivial factorization @xmath488 corresponds to the solution @xmath489 and @xmath490 .
i would like to thank john quintanilla and natasa jonoska for their useful discussions .
frobenius , ber quadratische formen , die viele primzahlen darstellen , " _ sitzungsber .
akad . wiss .
zu berlin _ ,
( 1912 ) , pp .
966 - 980 .
reprinted in _
gesammelte abhadlungen _ , vol .
iii , 573 - 587 .
springer - verlag , berlin .
( 1968 ) .
landau , e. , gelste und ungelste probleme aus der theorie der primzahlverteilung und der riemannschen zetafunktion , " _ proc . of the fifth internat .
congr . of math .
_ , cambridge , aug .
22 - 28 , 1912 , * 1 * ( 1913 ) , pp . | we identify a recursive structure among factorizations of polynomial values into two integer factors
. polynomials for which this recursive structure characterizes all non - trivial representations of integer factorizations of the polynomial values into two parts are here called recursively - factorable polynomials . in particular
, we prove that @xmath0 and the prime - producing polynomials @xmath1 and @xmath2 are recursively - factorable . for quadratics ,
the we prove that this recursive structure is equivalent to a diophantine identity involving the product of two binary quadratic forms .
we show that this identity may be transformed into geometric terms , relating each integer factorization @xmath3 to a lattice point of the conic section @xmath4 , and vice versa . |
stars are formed inside molecular cloud cores by means of gravitational collapse .
the details of the process , however , are complicated by the presence of magnetic fields and angular momentum . as a result ,
in addition to infall ( or collapse ) , rotation and outflow are also seen toward star - forming regions . in theory ,
a rotationally supported disk is expected to form inside a collapsing core around a protostar , from which part of the material is accreted by the protostar and part is ejected away .
observationally , however , when and how a rotationally supported disk is actually formed are still unclear , because of the lack of detailed kinematic studies inside the collapsing core .
a rotationally supported disk has been seen with a radius of @xmath0 500 au in the late ( i.e. , class ii or t tauri ) phase of star formation ( see , e.g. , * ? ? ?
formation of such a disk must have started early in the class 0 phase . however , no such disk has been confirmed in the class 0 phase , probably because it is still too small ( @xmath1 100 au ) for a detailed kinematic study with current instruments ( see , e.g. , * ? ? ?
* ; * ? ? ?
in order to investigate how a rotationally supported disk is formed inside a collapsing core , we have mapped the protostellar system hh 111 , which is in the class i phase and is thus expected to have a bigger disk than that in the class 0 phase . in this system ,
a class i source iras 05491 + 0247 is found deeply embedded in a compact molecular cloud core in the l1617 cloud of orion , driving the hh 111 jet @xcite .
the distance to the orion nebula cluster has been recently refined to be 400 pc @xcite , the same that is generally adopted for the orion b cloud ( see , e.g. , * ? ? ?
* ) with which l1617 is associated .
the hh 111 system has traditionally been assumed to be at a distance of 450 pc , but we here adopt this new distance , at which the source has a luminosity of @xmath0 20 @xcite .
a rotating envelope has been seen around the source inside the core @xcite and may have infall motion toward the source @xcite .
the jet complex is highly collimated with a two - sided length of @xmath0 6.7 pc ( * ? ? ?
* corrected for the new distance ) , and it produces a collimated molecular outflow around it @xcite .
the source has also been detected with the vla , associated with a tiny radio jet @xcite , and a tiny dusty disk ( with a radius of 30 au ) @xcite that can be the innermost part of a rotationally supported disk .
thus , this system is one of the best candidates to search for a rotationally supported disk and investigate how it is formed inside a collapsing core . here
, we present observations of the envelope in 1.33 mm continuum , ( @xmath2 ) , ( @xmath2 ) , and so ( @xmath3 ) emission obtained with the submillimeter array ( sma ) @xcite .
observations toward the protostellar system hh 111 were carried out with the sma on 2005 december 5 in the compact - north configuration using the 230 ghz band receivers .
the receivers have two sidebands , lower and upper . with the local oscillator ( lo )
frequency set to 224.677 ghz , we observed the ( @xmath2 ) line in the upper sideband , the ( @xmath2 ) , ( @xmath2 ) , and so ( ) lines in the lower sideband simultaneously .
the line traces mainly the outflow interaction and will be presented in a future publication . combining the line - free portions of the two sidebands results in a total continuum bandwidth of @xmath0 3.9 ghz centered at the lo frequency ( or @xmath4 1.33 mm ) for the continuum .
seven antennas were used , giving baselines with projected lengths ranging from 8 to 53 m. the correlator was set up to give a velocity resolution of 0.28 per channel for the , , and lines , and 0.56 per channel for the so line .
three pointings with a separation of @xmath0 were used to map the envelope and the outflow lobe extending to the west . here , we only present the envelope region close to the source .
the visibility data were calibrated with the mir package , with uranus as a flux calibrator , quasar 3c454.3 as a passband calibrator , and quasar 0530 + 135 as a gain calibrator .
the flux uncertainty is estimated to be @xmath0 15 % .
the calibrated visibility data were imaged with the miriad package .
the dirty maps that were produced from the calibrated visibility data were cleaned using the steer clean method , producing the clean component maps .
the final maps were obtained by restoring the clean component maps with a synthesized ( gaussian ) beam fitted to the main lobe of the dirty beam . with natural weighting ,
the synthesized beam has a size of @xmath5 at a position angle ( p.a . ) of @xmath0 @xmath650 .
the rms noise levels are @xmath0 0.22 in the and channel maps , 0.15 in the so channel maps , and 3.5 in the continuum map .
the velocities of the channel maps are lsr .
our results are presented in comparison to a mosaic image based on the hubble space telescope ( hst ) nicmos images ( [ feii ] 1.64 + 2 at 2.12 + continuum ) obtained by @xcite , which shows clear detection of two infrared sources , reflection nebulae tracing the illuminated outflow cavity walls , and the jet in the system .
the two infrared sources are named stars a and b , with star a being the iras source .
they are also detected in 3.6 cm continuum by the vla as the vla 1 and 2 sources , respectively , at @xmath7 , @xmath8 and @xmath9 , @xmath10 @xcite .
these vla positions , which are more accurate than the nicmos positions , are adopted here as the source positions . in this region ,
the systemic velocity is assumed to be given by the line peak of the optically thin so emission , which is [email protected] lsr .
note that this value is higher than that derived from the emission averaged over a much larger region by @xcite , which was 8.5 . throughout this paper ,
the velocity is relative to this systemic value .
a compact continuum emission is detected at 1.33 mm , with an emission peak well coincident with the vla 1 position ( fig .
[ fig : cont]a ) .
it has a total flux of 280@xmath1140 mjy , about 60% of that measured for a much larger region in the single - dish observation @xcite .
the emission is spatially unresolved , mainly arising from around the vla 1 source .
some faint emission is also seen extending towards the vla 2 source . from the spectral energy distribution ( sed ) ( fig .
[ fig : cont]b ) , it is clear that the emission at 1.33 mm is thermal emission from the dust in the envelope and the putative disks around the vla 1 and 2 sources @xcite .
assuming that the mass opacity is given by @xmath12 the sed between @xmath133.06 mm and 25 can be fitted with @xmath14 and a dust temperature of 41@xmath664 k. with these fits , the emission is found to be optically thin at 1.33 mm .
the total mass of the gas and dust associated with the envelope and the putative disks is estimated to be @xmath0 [email protected] . unlike the continuum emission , , , and
so emission are all seen with a peak near the center of the dark ridge ( i.e. , the dust lane ) seen in the hst image , which is @xmath0 to the west of the vla 1 source ( fig .
[ fig : linemap ] ) .
the emission shows similar structures to those seen in cs @xcite .
it shows a north - south elongation roughly perpendicular to the jet axis , tracing a flattened envelope around the vla 1 source .
the emission also extends slightly to the west along the jet axis and southwest along the outflow cavity wall outlined by the reflection nebulae .
the emission is more extended , tracing both the envelope and the outflow cavity walls . the so emission is more compact with a narrow waist around the source .
it also extends along the outflow cavity walls in the southeast , southwest , and northwest directions .
the dark ridge , where the line emissions all peak , has been suggested to be the birthplace of both the vla 1 and 2 sources @xcite and thus can be assumed to be the center of the flattened envelope .
the , , and so spectra toward the dark ridge all show a similar fwzi of @xmath0 6 ( fig .
[ fig : spec ] ) .
the so spectrum is rather symmetric with a single peak at the systemic velocity . the and spectra show a double - peaked profile with a dip coincident with the so peak , and with the blueshifted peak brighter than the redshifted peak ( i.e. , the blue asymmetry ) . comparing to the single - dish observations toward a larger region
, we find that there is missing flux in and within @xmath0 1 of the systemic velocity as the large - scale flux is resolved out in our interferometric observation ( fig .
[ fig : spec_sd ] ) .
note that , however , since missing flux tends to produce a dip with similar redshifted peak and blueshifted peak , an additional mechanism is still needed to produce the observed dips with the blue asymmetry . it could be self - absorption due to an infall motion in the envelope ( see , e.g. , * ? ? ?
infall motion has already been suggested by @xcite using cs observations .
the emission is optically thicker and thus shows a deeper and wider absorbing dip than the emission .
the so emission does not show a dip , likely because it is optically thin due to low so abundance . in order to show the kinematics of the flattened envelope ,
the emission is divided into four velocity components : low redshifted and blueshifted emission with velocity within @xmath0 1 from the systemic , and high redshifted and blueshifted emission with velocity higher than 1 from the systemic ( fig .
[ fig : linemap ] ) . at low velocity
, the emission clearly shows a flattened structure with the blueshifted and redshifted emission extended @xmath0 ( 2800 au ) to the north and south , respectively , from the dark ridge , as expected for a rotating envelope .
the redshifted emission fits right in between the eastern and western reflection nebulae in the south .
this is to be expected since the reflection nebulae trace the regions with less extinction and thus low density in the boundary of the envelope .
therefore , with the boundary defined by those reflection nebulae , the envelope is found to have a thickness of @xmath0 ( i.e. , 1000 au ) and roughly constant with increasing distance from the source .
note that the blueshifted emission also traces the outflow emission in the west and the outflow cavity wall in the southwest . at high velocity
, the emission shrinks toward to the source , arising from the inner part of the envelope .
the blueshifted peak shifts slightly to the west and the redshifted peak slightly to the east .
these shifts further suggest an infall motion in the envelope , with the blueshifted peak in the far end ( west ) of the envelope and the redshifted peak in the near end ( east ) , as suggested in @xcite .
these shifts are also partly due to an outflow component , with the emission extended to the west and along the cavity walls . in
, the blueshifted emission peaks to the northwest and redshifted emission in the southeast of the dark ridge even at low velocity .
as seen in the spectrum , the lowest - velocity emission , which traces the flattened envelope further out , is missing more than the emission , and thus revealing the inner part of the envelope , where the radial motion becomes prominent .
in so , the blueshifted and redshifted emission are also seen to the northwest and southwest of the dark ridge , similar to that seen in .
the so emission is likely optically thin , probing directly the inner part of the envelope .
the rotational velocity of the flattened envelope can be studied with the position - velocity ( pv ) diagrams centered at the dark ridge cut perpendicular to the jet axis .
the and emission show similar kinematics ( figs .
[ fig : pvs]a , b ) and can be studied together .
the blueshifted emission is seen mainly to the north and redshifted emission mainly to the south , confirming a rotation in the envelope . at low velocity ,
the blueshifted and redshifted emissions also extend across the center to the south and north , respectively , suggesting that there is also a radial motion in the envelope .
the rotational velocity is seen increasing toward the center and thus could be keplerian or that conserving specific angular momentum . in the case of keplerian rotation ,
the best fit to the velocity structures in these two line emissions together is @xmath15 with @xmath16 au ( i.e. , ) , which is slightly faster than that found in cs at lower angular resolution by @xcite .
this keplerian rotation corresponds to a stellar mass of @xmath0 0.8 . here
, the plane of the flattened envelope is assumed to have an inclination of 10 away from the line of sight @xcite . in the case of conservation of specific angular momentum ,
the best fit is @xmath17 . with a reduced @xmath18
, it seems to fit better than the keplerian rotation , which has a reduced @xmath19 and decreases too fast toward the center as compared with the observed . to confirm it , however ,
both observations at higher angular resolution and shorter @xmath20 spacings are needed .
the latter is to retrieve the missing flux at low velocity .
the so emission traces only the inner part of the envelope .
its velocity structure is not well resolved and seems to have two components : a low - velocity component in the inner part with the velocity within @xmath0 1.5 of the systemic , and a high - velocity component in the outer part with the velocity higher than @xmath0 1.5 ( fig .
[ fig : pvs]c ) .
the low - velocity component seems to have a radius of @xmath0 ( 200 au ) and a differential rotation with the velocity increasing toward the center .
the high - velocity component extends to @xmath21 1.3@xmath22 ( @xmath0 500 au ) , connecting to that of the emission .
its velocity , however , seems to increase with increasing distance from the center .
it could arise from a solid - body ( rigid ) rotation with @xmath23 .
it could also arise from an outflow motion .
the radial motion in the envelope can be studied with the pv diagrams cut along the jet axis ( figs .
[ fig : pvs]d , e , f ) .
the pv diagrams , however , are highly contaminated by the outflow emission , especially on the blueshifted side
. only part of the emission close to the center to within e.g. , ( which corresponds to a deprojected distance of @xmath0 at an inclination of @xmath0 10 ) is believed to be from the envelope .
the , , and so emission within @xmath0 from the center seem to show similar velocity structure .
the blueshifted emission is mainly in the west and the redshifted emission is mainly in the east . on the redshifted side ,
where the outflow contamination is less , the velocity seems to increase toward the center .
this velocity structure is unlikely due to an outflow motion for which the velocity is expected to decrease toward the center ( see , e.g , * ? ? ?
it could arise from an infall motion in the envelope , as suggested . for a dynamical collapse
, the infall motion can be assumed to be a free - fall motion as in @xcite , with the infall velocity @xmath24 .
three values of @xmath25 , 2.4 , 1.9 , and 1.5 , are compared to the , , and so velocity structures , with the second value corresponding to that obtained by @xcite .
the corresponding stellar masses are 1.3 , 0.8 , 0.5 , respectively .
the one with the least mass seems to give the best fit , matching the emission at the lowest velocity in the redshifted part where the contamination of the outflow emission is less .
however , observation at higher angular resolution is really needed to confirm it . in order to obtain roughly the density distribution of the flattened envelope , a simple disklike envelope model as described in @xcite is compared to the emission , which suffers less missing flux than the emission . in this model , the envelope has a free - fall motion and a rotation with conservation of specific angular momentum , with the best - fit parameters given above .
it has an inner radius of @xmath26 , an outer radius of @xmath27 , and a constant thickness of @xmath28 , in cylindrical coordinates .
the number density of molecular hydrogen is assumed to be given by @xmath29 where @xmath30 is the density at @xmath31 and @xmath32 is a power - law index assumed to be @xmath33 , as in many theoretical infalling models ( see , e.g. , * ? ? ?
* ; * ? ? ?
the temperature of the envelope is uncertain and assumed to be given by @xmath34 where @xmath35 is the temperature at @xmath31 and @xmath36 is a power - law index .
the abundance of relative to molecular hydrogen is assumed to be constant and given by @xmath37 @xcite . in the model calculations ,
radiative transfer is used to calculate the emission , with an assumption of local thermal equilibrium .
for simplicity , the line width is assumed to be given by the thermal line width only .
the channel maps of the emission derived from the model are convolved with the observed beam and velocity resolution , and then used to make the integrated map , spectrum , and pv diagram .
there are six parameters in this model : @xmath26 , @xmath27 , @xmath28 , @xmath30 , @xmath35 , and @xmath36 . in our fits
, we have @xmath38 au ( @xmath0 ) , @xmath39 au ( @xmath0 ) , @xmath40 1000 au ( i.e. , ) , @xmath41 , @xmath42 k , and @xmath43 . here
@xmath44 is assumed to be the lower end of the dust temperature . as can be seen in fig .
[ fig : modelc18o ] , this model can roughly reproduce the envelope structure , spectrum , and pv diagram of the emission , except for the emission at high - redshifted velocity . the position of the two peaks in the line profile can also be roughly matched with the assumed infall velocity .
note that , however , the temperature power - law index is required to be steeper here in order to reproduce the observed dip at the systemic velocity , as compared to those ( @xmath45 ) obtained by fitting other envelopes in single - dish observations ( see , e.g. , * ? ? ?
. it could be because the dip is partly due to the missing flux of the large - scale gas and the absorption by this large - scale gas as well .
the mass of the flattened envelope , which is given by @xmath46 is @xmath47 , consistent with that found by @xcite .
this mass is higher than that derived from the continuum emission , which is more compact . in our simple model
, the infall rate , which is given by @xmath48 has a mean value of @xmath49 , averaged over the radius .
the accretion rate , if assumed to be the mean infall rate , would result in an accretion luminosity of @xmath50 , assuming a stellar mass of @xmath51 ( as derived from the infall velocity ) and a stellar radius of @xmath52 .
this luminosity , however , is about twice the luminosity of the source , which is @xmath0 20 @xcite .
in addition , the accretion time would be only @xmath0 5@xmath53 yr for forming a source with a mass of 0.5 , shorter than that for a typical class i source .
thus , the accretion rate could be smaller than the infall rate .
the envelope seen in and can be considered as a thick torus around the narrow dust lane ( i.e. , the dark ridge ) , with an inner radius of 400 au ( ) , an outer radius of 3200 au ( ) , and a thickness of 1000 au ( ) .
it has a differential rotation that seems to be better fitted by a rotation with conservation of specific angular momentum than by a keplerian rotation as assumed by @xcite .
the envelope likely also has an infall motion , with the blueshifted emission in the far end and the redshifted emission in the near end , showing a double - peaked line profile with a blue asymmetry and an absorption dip at the systemic velocity .
therefore , the envelope seems to be infalling ( contracting ) toward the center preserving its angular momentum , similar to other infalling envelopes in the class i phase @xcite and even in the class 0 phase ( e.g. , * ? ? ?
* ) . with the derived rotation and infall velocity structures , the rotation velocity is equal to the infall velocity at @xmath54 ( 600 au ) , suggesting that the centrifugal force starts to dominate in the inner edge of the envelope .
the inner envelope is seen in so as in hh 212 @xcite .
it has an outer radius similar to where the centrifugal force starts to dominate .
therefore , a rotationally supported disk is expected to form here .
such a disk may indeed have formed in the inner part of this inner envelope , which is seen with a differential rotation , spatially coincident with the narrow dark ridge that could be interpreted as a dense circumstellar disk @xcite . the outer part of this inner envelope , however , may have a rotation velocity decreasing toward the center .
if this is the case , the angular momentum might have been efficiently removed from there .
this can be achieved with the presence of a magnetic field .
it has been found that a magnetic field can enforce solid - body rotation because magnetic braking can eliminate differential rotation on a timescale much shorter than even the free - fall timescale @xcite .
currently , the most popular model of star formation is shu s model ( derived from * ? ? ?
one recent version of this model assumes the collapse of a magnetized singular isothermal rotating toroid @xcite . in this model
, a pseudodisk ( a nonrotationally supported thick torus ) is formed around the source . in the inner edge of the pseudodisk ,
the rotation speed drops steadily with decreasing radius , as a result of efficient braking by the strong magnetic field . in our observations
, it is possible that the flattened envelope seen in and is a pseudodisk resulting from the collapse of a magnetized rotating toroid toward the center .
it is also possible that the outer part of the so envelope is the inner edge of the pseudodisk where the rotation velocity decreases with decreasing radius due to efficient magnetic braking .
however , further observations are really needed to confirm the rotation velocity structure in the flattened envelope and to tell whether the structure of the infalling material is magnetically - guided .
in addition , a slow outflow , which is expected when there is a twisting of magnetic field lines @xcite , has not yet been identified in our observations . in their model ,
a rotationally supported disk is not able to form around the source due to the efficient loss of angular momentum @xcite . in the hh 111 system
, such a disk must have already formed around the source to launch the jet .
in addition , the innermost part of the disk has been seen as a tiny dusty disk ( with a radius of 30 au ) around the source @xcite . as suggested by @xcite ,
one possibility is to have an ohmic resistivity in the model , so that the magnetic field can be dissipated efficiently around the source .
the continuum emission peaks at the vla 1 position while the line emissions all peak around the dark ridge , suggesting that the dust and gas peak at different positions .
it could be because there is a lack of gas toward the continuum peak .
it has been suggested by @xcite that the vla 1 and vla 2 sources were originally formed in the dark ridge inside the envelope but later ejected from there .
thus , the continuum emission peak could trace the inner disk carried away by the vla 1 source , as suggested in @xcite .
if this is the case , there could be a lack of gas ( depletion ) toward the inner disk .
we have mapped the central region of the protostellar system hh 111 in 1.33 mm continuum , ( j=2 - 1 ) , ( j=2 - 1 ) , and so ( ) emission at @xmath0 resolution with the submillimeter array .
there are two sources , vla 1 ( = iras 05491 + 0247 ) and vla 2 , with the vla 1 source driving the hh 111 jet .
thermal emission is seen in 1.33 mm continuum tracing the dust in the envelope and the putative disks around the sources . a flattened
, torus - like envelope is seen in and around the vla 1 source surrounding the dust lane perpendicular to the jet axis , with an inner radius of @xmath0 400 au ( ) , an outer radius of @xmath0 3200 au ( ) , and a thickness of @xmath0 1000 au ( ) .
it seems to be infalling toward the center with conservation of specific angular momentum rather than with a keplerian rotation as assumed by @xcite .
an inner envelope is seen in so , with a radius of @xmath0 500 au ( ) . the inner part of this inner envelope , which is spatially coincident with the dust lane , seems to have a differential rotation and thus may have formed a rotationally supported disk .
the outer part of this inner envelope , however , may have a rotation velocity decreasing toward the center and thus represent a region where an infalling envelope is in transition to a rotationally supported disk .
a brief comparison with a collapsing model suggests that the flattened , torus - like envelope seen in and could be a pseudodisk ( a nonrotationally supported torus ) resulting from a collapse of a magnetized rotating toroid .
we thank the sma staff for their efforts in running and maintaining the array .
we also thank the referee , dan watson , for his valuable comments .
br is partly supported by the nasa astrobiology institute under cooperative agreement no .
nna04cc08a issued through the office of space science .
allen , a. , li , z. , & shu , f. h. 2003 , , 599 , 363 cernicharo , j. , & reipurth , b. 1996 , , 460 , l57 evans , n. j. , ii 1999 , , 37 , 311 frerking , m. a. , langer , w. d. , & wilson , r. w. 1982 , , 262 , 590 gibb , a.g .
2008 , in handbook of star forming regions , ed .
bo reipurth , astron .
pacific , in press ho , p. t. p. , moran , j. m. , & lo , k. y. 2004 , , 616 , l1 hogerheijde , m. r. 2001 , , 553 , 618 jrgensen , j. k. , et al .
2007 , , 659 , 479 lee , c .- f . , ho , p. t. p. , beuther , h. , bourke , t. l. , zhang , q. , hirano , n. , & shang , h. 2006 , , 639 , 292 lee , c .- f . , mundy , l. g. , reipurth , b. , ostriker , e. c. , & stone , j. m. 2000 , , 542 , 925 lee , c .- f . , stone , j. m. , ostriker , e. c. , & mundy , l. g. 2001 , , 557 , 429 lee , c .- f . , & ho , p. t. p. 2005
, , 632 , 964 lee , c .- f . , ho , p. t. p. , & white , s. m. 2005 , , 619 , 948 mellon , r. r. , & li , z .- y . 2008 , , 681 , 1356 menten , k. m. , reid , m. j. , forbrich , j. , & brunthaler , a. 2007 , , 474 , 515 momose , m. , ohashi , n. , kawabe , r. , nakano , t. , & hayashi , m. 1998 , , 504 , 314 mouschovias , t. c. & ciolek , g. e. 1999 , nato asic proc . 540 : the origin of stars and planetary systems , 305 nagar , n. m. , vogel , s. n. , stone , j. m. , & ostriker , e. c. 1997 , , 482 , l195 nakamura , f. 2000 , , 543 , 291 ohashi , n. , hayashi , m. , ho , p. t. p. , momose , m. , tamura , m. , hirano , n. , & sargent , a. i. 1997 , , 488 , 317 reipurth , b. , chini , r. , krugel , e. , kreysa , e. , & sievers , a. 1993 , , 273 , 221 reipurth , b. 1989 , , 340 , 42 reipurth , b. , bally , j. , & devine , d. 1997 , , 114 , 2708 reipurth , b. , & olberg , m. 1991 , , 246 , 535 reipurth , b. , raga , a. c. , & heathcote , s. 1992 , , 392 , 145 reipurth , b. , yu , k. c. , rodrguez , l. f. , heathcote , s. , & bally , j. 1999 , , 352 , l83 rodrguez , l. f. , torrelles , j. m. , anglada , g. , & reipurth , b. 2008 , , 136 , 1852 sandstrom , k. m. , peek , j. e. g. , bower , g. c. , bolatto , a. d. , & plambeck , r. l. 2007 , , 667 , 1161 shu , f. h. , galli , d. , lizano , s. , & cai , m. 2006 , , 647 , 382 simon , m. , dutrey , a. , & guilloteau , s. 2000 , , 545 , 1034 stapelfeldt , k. r. , & scoville , n. z. 1993 , , 408 , 239 shu , f. h. 1977 , , 214 , 488 yang , j. , ohashi , n. , yan , j. , liu , c. , kaifu , n. , & kimura , h. 1997 , , 475 , 683 | we have mapped the central region of the hh 111 protostellar system in 1.33 mm continuum , ( j=2 - 1 ) , ( j=2 - 1 ) , and so ( ) emission at @xmath0 resolution with the submillimeter array .
there are two sources , vla 1 ( = iras 05491 + 0247 ) and vla 2 , with the vla 1 source driving the hh 111 jet .
thermal emission is seen in 1.33 mm continuum tracing the dust in the envelope and the putative disks around the sources . a flattened
, torus - like envelope is seen in and around the vla 1 source surrounding the dust lane perpendicular to the jet axis , with an inner radius of @xmath0 400 au ( ) , an outer radius of @xmath0 3200 au ( ) , and a thickness of @xmath0 1000 au ( ) .
it seems to be infalling toward the center with conservation of specific angular momentum rather than with a keplerian rotation as assumed by @xcite .
an inner envelope is seen in so , with a radius of @xmath0 500 au ( ) . the inner part of this inner envelope , which is spatially coincident with the dust lane , seems to have a differential rotation and thus may have formed a rotationally supported disk .
the outer part of this inner envelope , however , may have a rotation velocity decreasing toward the center and thus represent a region where an infalling envelope is in transition to a rotationally supported disk .
a brief comparison with a collapsing model suggests that the flattened , torus - like envelope seen in and could result from a collapse of a magnetized rotating toroid . |
over the past ten years , the study of phase transition phenomena has been one of the most exciting areas in computer science and artificial intelligence .
numerous studies have established that for many np - complete problems ( e.g. , sat and csp ) , the hardest random instances occur , while a control parameter is varied accordingly , between an under - constrained region where all instances are almost surely satisfiable and an over - constrained region where all instances are almost surely unsatisfiable . in the transition region , there is a threshold where half the instances are satisfiable . generating hard instances is important both for understanding the complexity of the problems and for providing challenging benchmarks @xcite .
another remarkable progress in artificial intelligence has been the development of incomplete algorithms for various kinds of problems . and
, since this progress , one important issue has been to produce hard satisfiable instances in order to evaluate the efficiency of such algorithms , as the approach that involves exploiting a complete algorithm in order to keep random satisfiable instances generated at the threshold can only be used for instances of limited size . also , it has been shown that generating hard ( forced ) satisfiable instances is related to some open problems in cryptography such as computing a one - way function @xcite . in this paper
, we mainly focus on random csp ( constraint satisfaction problem ) instances .
initially , four `` standard '' models , denoted a , b , c and d @xcite , have been introduced to generate random binary csp instances .
however , @xcite have identified a shortcoming of all these models .
indeed , they prove that random instances generated using these models suffer from ( trivial ) unsatisfiability as the number of variables increases . to overcome the deficiency of these standard models ,
several alternatives have been proposed . on the one hand ,
@xcite have proposed a model e and @xcite a generalized model .
however , model e does not permit to tune the density of the instances and the generalized model requires an awkward exploitation of probability distributions .
also , other alternatives correspond to incorporating some `` structure '' in the generated random instances .
roughly speaking , it involves ensuring that the generated instances be arc consistent @xcite or path consistent @xcite .
the main drawback of all these approaches is that generating random instances is no more quite a natural and easy task . on the other hand ,
@xcite , @xcite and @xcite have revisited standard models by controlling the way parameters change as the problem size increases .
the alternative model d scheme of @xcite guarantees the occurrence of a phase transition when some parameters are controlled and when the constraint tightness is within a certain range .
the two revised models , called rb and rd , of @xcite provide the same guarantee by varying one of two control parameters around a critical value that , in addition , can be computed . also , @xcite identify a range of suitable parameter settings in order to exhibit a non - trivial threshold of satisfiability .
their theoretical results apply to binary instances taken from model a and to `` symmetric '' binary instances from a so - called model b which , not corresponding to the standard one , associates the same relation with every constraint . the models rb and rd present several nice features : * it is quite easy to generate random instances of any arity as no particular structure has to be integrated , or property enforced , in such instances . *
the existence of an asymptotic phase transition can be guaranteed while applying a limited restriction on domain size and on constraint tightness . for instances involving constraints of arity @xmath0
, the domain size is required to be greater than the k@xmath1 root of the number of variables and the ( threshold value of the ) constraint tightness is required to be at most @xmath2 .
* when the asymptotic phase transition exists , a threshold point can be precisely located , and all instances generated following models rb and rd have the guarantee to be hard at the threshold , i.e. , to have an exponential tree - resolution complexity . *
it is possible to generate forced satisfiable instances whose hardness is similar to unforced satisfiable ones .
this paper is organized as follows . after introducing models rb and rd , as well as some theoretical results ( section [ sec : the ] ) ,
we provide a formal analysis about generating both forced and unforced hard satisfiable instances ( section [ sec : generating ] ) .
then , we present the results of a large series of experiments that we have conducted ( section [ sec : experimental ] ) , and , before concluding , we discuss some related work ( section [ sec : related ] ) .
a constraint network consists of a finite set of variables such that each variable @xmath3 has an associated domain denoting the set of values allowed for @xmath3 , and a a finite set of constraints such that each constraint @xmath4 has an associated relation denoting the set of tuples allowed for the variables involved in @xmath4 .
a solution is an assignment of values to all the variables such that all the constraints are satisfied .
a constraint network is said to be satisfiable ( sat , for short ) if it admits at least a solution .
the constraint satisfaction problem ( csp ) , whose task is to determine if a given constraint network , also called csp instance , is satisfiable , is np - complete . in this section ,
we introduce some theoretical results taken from @xcite .
first , we introduce a model , denoted rb , that represents an alternative to model b. note that , unlike model b , model rb allows selecting constraints with repetition .
but the main difference of model rb with respect to model b is that the domain size of each variable grows polynomially with the number of variables .
a class of random csp instances of model rb is denoted rb(@xmath0,@xmath5,@xmath6,@xmath7,@xmath8 ) where : * @xmath9 denotes the arity of each constraint , * @xmath5 denotes the number of variables , * @xmath10 determines the domain size @xmath11 of each variable , * @xmath12 determines the number @xmath13 of constraints , * @xmath14 denotes the tightness of each constraint . to build one instance @xmath15 rb(@xmath0,@xmath5,@xmath6,@xmath7,@xmath8 ) , we select with repetition @xmath16 constraints , each one formed by selecting @xmath0 distinct variables and @xmath17 distinct unallowed tuples ( as @xmath8 denotes a proportion ) . when fixed , @xmath6 and @xmath7 give an indication about the growth of the domain sizes and of the number of constraints as @xmath5 increases since @xmath11 and @xmath18 , respectively .
it is then possible , for example , to determine the critical value @xmath19 of @xmath8 where the hardest instances must occur .
indeed , we have @xmath20 which is equivalent to the expression of @xmath19 given by @xcite .
another model , denoted model rd , is similar to model rb except that @xmath8 denotes a probability instead of a proportion . for convenience , in this paper , we will exclusively refer to model rb although all given results hold for both models . in @xcite , it is proved that model rb , under certain conditions , not only avoids trivial asymptotic behaviors but also guarantees exact phase transitions .
more precisely , with pr denoting a probability distribution , the following theorems hold .
[ the:1 ] if @xmath0 , @xmath21 and @xmath22 are constants then @xmath23 = \left \ { \begin{array}{cc } 1 & if~ r < r_{cr } \\ 0 & if~ r > r_{cr } \end{array } \right.\end{aligned}\ ] ] where @xmath24 .
[ the:2 ] if @xmath0 , @xmath21 and @xmath25 are constants then @xmath23 = \left \ { \begin{array}{cc } 1 & if~ p < p_{cr } \\ 0 & if~ p > p_{cr } \end{array } \right.\end{aligned}\ ] ] where @xmath26 .
remark that the condition @xmath25 is equivalent to @xmath27 given in @xcite .
theorems [ the:1 ] and [ the:2 ] indicate that a phase transition is guaranteed provided that the domain size is not too small and the constraint tightness or the threshold value of the constraint tightness not too large . as an illustration , for instances involving binary ( resp .
ternary ) constraints , the domain size is required to be greater than the square ( resp .
cubic ) root of the number of variables and the constraint tightness or threshold value of the tightness is required to be at most @xmath28 ( resp .
@xmath29 ) .
the following theorem establishes that unsatisfiable instances of model rb almost surely have the guarantee to be hard . a similar result for model
a has been obtained by @xcite with respect to binary instances .
[ the:3 ] if @xmath15 rb(@xmath0,@xmath5,@xmath6,@xmath7,@xmath8 ) and @xmath0 , @xmath6 , @xmath7 and @xmath8 are constants , then , almost surely as the number of variables tends to infinity . ]
, @xmath30 has no tree - like resolution of length less than @xmath31 .
the proof , which is based on a strategy following some results of @xcite , is omitted but can be found in @xcite . to summarize , model rb guarantees exact phase transitions and hard instances at the threshold .
it then contradicts the statement of @xcite about the requirement of an extremely low tightness for all existing random models in order to have non - trivial threshold behaviors and guaranteed hard instances at the threshold .
for csp and sat , there is a natural strategy to generate _ forced _
satisfiable instances , i.e. , instances on which a solution is imposed .
it suffices to proceed as follows : first generate a random ( total ) assignment @xmath32 and then generate a random instance with @xmath5 variables and @xmath16 constraints ( clauses for sat ) such that any constraint violating @xmath32 is rejected .
@xmath32 is then a _ forced _ solution .
this strategy , quite simple and easy to implement , allows generating hard forced satisfiable instances of model rb provided that theorem @xmath33 or @xmath34 holds .
nevertheless , this statement deserves a theoretical analysis .
assuming that @xmath35 denotes the domain size ( @xmath35 = @xmath34 for sat ) , we have exactly @xmath36 possible ( total ) assignments , denoted by @xmath37 , and @xmath38 possible assignment pairs where an _ assignment pair _
@xmath39 is an ordered pair of two assignments @xmath40 and @xmath41 .
we say that @xmath39 satisfies an instance if and only if both @xmath40 and @xmath41 satisfy the instance .
then , the expected ( mean ) number of solutions @xmath42 $ ] for instances that are forced to satisfy an assignment @xmath40 is : @xmath43 = \overset{d^n}{\underset{j=1}{\sum } } \frac{\pr[<t_i , t_j>]}{\pr[<t_i , t_i>]}\ ] ] where @xmath44 $ ] denotes the probability that @xmath39 satisfies a random instance .
note that @xmath42 $ ] should be independent of the choice of the forced solution @xmath40 .
so we have : @xmath43=\frac{\underset{1\leq i , j\leq d^{n}}{{\displaystyle\sum } } \pr[<t_{i},t_{j}>]}{d^{n}\pr[<t_{i},t_{i}>]}=\frac{e[n^{2}]}{e[n]}.\ ] ] where @xmath45 $ ] and @xmath46 $ ] are , respectively , the second moment and the first moment of the number of solutions for random unforced instances . for random 3-sat
, it is known that the strategy mentioned above is unsuitable as it produces a biased sampling of instances with many solutions clustered around @xmath32 @xcite .
experiments show that forced satisfiable instances are much easier to solve than unforced satisfiable instances .
in fact , it is not hard to show that , asymptotically , @xmath45 $ ] is exponentially greater than @xmath47 $ ] .
thus , the expected number of solutions for forced satisfiable instances is exponentially larger than the one for unforced satisfiable instances .
it then gives a good theoretical explanation of why , for random 3-sat , the strategy is highly biased towards generating instances with many solutions .
for model rb , recall that when the exact phase transitions were established @xcite , it was proved that @xmath45/e^{2}[n]$ ] is asymptotically equal to @xmath33 below the threshold , where almost all instances are satisfiable , i.e. @xmath45/e^{2}[n]\approx 1 $ ] for @xmath48 or @xmath49 .
hence , the expected number of solutions for forced satisfiable instances below the threshold is asymptotically equal to the one for unforced satisfiable instances , i.e. @xmath42=e[n^{2}]/e[n]\approx e[n]$ ] . in other words , when using model rb , the strategy has almost no effect on the number of solutions and does not lead to a biased sampling of instances with many solutions .
in addition to the analysis above , we can also study the influence of the strategy on the distribution of solutions with respect to the forced solution .
we first define the distance @xmath50 between two assignments @xmath40 and @xmath41 as the proportion of variables that have been assigned a different value in @xmath40 and @xmath41 .
we have @xmath51 .
for forced satisfiable instances of model rb , with @xmath52 $ ] denoting the expected number of solutions whose distance from the forced solution ( identified as @xmath40 , here ) is equal to @xmath53 , we obtain by an analysis similar to that in @xcite : @xmath54 = \overset{d^n}{\underset{j=1}{\sum } } \frac{\pr[<t_i , t_j>]}{\pr[<t_i , t_i>]}\text { \ \ with } d^{f}(t_i ,
t_j)=\delta\\ & = \binom{n}{n\delta } \left(n^{\alpha}{\text - } 1\right)^{n\delta } \left[\frac{\binom{n - n\delta}{k}}{\binom{n}{k}}+(1-p)\left ( 1-\frac{\binom{n - n\delta}{k}}{\binom{n}{k}}\right ) \right ] ^{rn\ln n}\\ & = \exp\left [ n\ln n\left ( r\ln\left(1-p+p(1-\delta)^{k}\right ) + \alpha \delta \right)+o(n ) \right ] .\end{aligned}\ ] ] it can be shown , from the results in @xcite that @xmath55 $ ] , for @xmath48 or @xmath49 , is asymptotically maximized when @xmath53 takes the largest possible value , i.e. @xmath56 . for unforced satisfiable instances of model rb , with @xmath57 $ ] denoting the expected number of solutions whose distance from @xmath40 ( not necessarily a solution ) is equal to @xmath53 , we have : @xmath58 & = \binom{n}{n\delta}\left ( n^{\alpha}-1\right ) ^{n\delta}\left ( 1-p\right ) ^{rn\ln n}\\ & = \exp\left [ n\ln n\left ( r\ln(1-p)+\alpha \delta \right)+o(n ) \right ] .\end{aligned}\ ] ] it is straightforward to see that the same pattern holds for this case , i.e. @xmath57 $ ] is asymptotically maximized when @xmath56 .
intuitively , with model rb , both unforced satisfiable instances and instances forced to satisfy an assignment @xmath32 are such that most of their solutions distribute far from @xmath32 .
this indicates that , for model rb , the strategy has little effect on the distribution of solutions , and is not biased towards generating instances with many solutions around the forced one . for random 3-sat , similarly
, we can show that as @xmath7 ( the ratio of clauses to variables ) approaches @xmath59 , @xmath55 $ ] and @xmath57 $ ] are asymptotically maximized when @xmath60 and @xmath61 , respectively .
this means , in contrast to model rb , that when @xmath7 is near the threshold , most solutions of forced instances distribute in a place much closer to the forced solution than solutions of unforced satisfiable instances .
as all introduced theoretical results hold when @xmath62 , the practical exploitation of these results is an issue that must be addressed . in this section , we give some representative experimental results which indicate that practice meets theory even if the number @xmath5 of variables is small . note that different values of parameters @xmath6 and @xmath7 have been selected in order to illustrate the broad spectrum of applicability of model rb . ,
@xmath7 and @xmath5,width=246 ] first , it is valuable to know in practice , to what extent , theorems [ the:1 ] and [ the:2 ] give precise thresholds according to different values of @xmath6 , @xmath7 and @xmath5 .
the experiments that we have run wrt theorem [ the:2 ] , as depicted in figure [ fig : diff ] , suggest that all other parameters being fixed , the greater the value of @xmath6 , @xmath7 or @xmath5 is , the more precise theorem [ the:2 ] is . more precisely , in figure [ fig : diff ] , the difference between the threshold theoretically located and the threshold experimentally determined is plotted against @xmath63 $ ] ( @xmath64 $ ] ) , against @xmath65 $ ] ( @xmath66 $ ] ) and against @xmath67 $ ] .
note that the vertical scale refers to the difference in constraint tightness and that the horizontal scale is normalized ( value @xmath68 respectively corresponds to @xmath69 , @xmath70 and @xmath71 , etc . ) .
instances ) of solving instances in rb(@xmath34,@xmath72,@xmath73,@xmath74,@xmath8),width=246 ] to solve the random instances generated by model rb , we have used a systematic backtracking search algorithm ( mac ) and a local search algorithm ( tabu search ) .
both algorithms have been equipped with a search heuristic that learns from conflicts @xcite .
we have studied the difficulty of solving with mac the binary instances of model rb generated around the theoretical threshold @xmath75 given by theorem [ the:2 ] for @xmath76 , @xmath77 , @xmath78 and @xmath79 . in figure
[ fig : phase1 ] , it clearly appears that the hardest instances are located quite close to the theoretical threshold and that the difficulty grows exponentially with @xmath5 ( note the use of a log scale ) .
it corresponds to a phase transition ( not depicted here , due to lack of space ) .
a similar behavior is observed in figure [ fig : phase2 ] with respect to ternary instances generated around the theoretical threshold @xmath80 for @xmath81 , @xmath82 , @xmath83 and @xmath84 .
instances ) of solving instances in rb(@xmath74,@xmath85,@xmath33,@xmath33,@xmath8),width=246 ] as the number and the distribution of solutions are the two most important factors determining the cost of solving satisfiable instances , we can expect , from the analysis given in section [ sec : generating ] , that for model rb , the hardness of solving forced satisfiable instances should be similar to that of solving unforced satisfiable ones .
this is what is observed in figure [ fig : phase1 ] . to confirm this ,
we have focused our attention to a point just below the threshold as we have then some ( asymptotic ) guarantee about the difficulty of both unforced and forced instances ( see theorems @xmath86 and @xmath87 in @xcite ) and the possibility of generating easily unforced satisfiable instances .
figure [ fig : forced ] shows the difficulty of solving with mac both forced and unforced instances of model rb at @xmath88 for @xmath76 , @xmath77 , @xmath89 and @xmath90 $ ] .
instances ) of solving instances in rb(@xmath34,@xmath91$],@xmath73,@xmath92,@xmath93),width=246 ] to confirm the inherent difficulty of the ( forced and unforced ) instances generated at the threshold , we have also studied the runtime distribution produced by a randomized search algorithm on distinct instances @xcite . for each instance
, we have performed @xmath94 independent runs .
figure [ fig : heavytail ] displays the survival function , which corresponds to the probability of a run taking more than @xmath95 backtracks , of a randomized mac algorithm for two representative instances generated at @xmath96 for @xmath76 , @xmath77 , @xmath89 and @xmath97 .
one can observe that the runtime distribution ( a log - log scale is used ) do not correspond to an heavy - tailed one , i.e. , a distribution characterized by an extremely long tail with some infinite moment .
it means that all runs behave homogeneously and , therefore , it suggests that the instances are inherently hard @xcite .
, @xmath98,@xmath73,@xmath92,@xmath96),width=246 ] then , we have focused on unforced unsatisfiable instances of model rb as theorem [ the:3 ] indicates that such instances have an exponential resolution complexity .
we have generated unforced and forced instances with different constraint tightness @xmath8 above the threshold @xmath96 for @xmath76 , @xmath77 , @xmath89 and @xmath99 $ ] .
figure [ fig : above ] displays the search effort of a mac algorithm to solve such instances against the number of variables @xmath5 .
it is interesting to note that the search effort grows exponentially with @xmath5 , even if the exponent decreases as the tightness increases .
also , although not currently supported by any theoretical result ( theorems @xmath86 and @xmath87 of @xcite hold only for forced instances below the threshold ) it appears here that forced and unforced instances have a similar hardness . instances ) of solving instances in rb(@xmath34,@xmath100$],@xmath73,@xmath92,@xmath8),width=246 ] finally , figure [ fig : tabu ] shows the results obtained with a tabu search with respect to the binary instances that have been previously considered with mac ( see figure [ fig : phase1 ] ) .
the search effort is given by a median cost since when using an incomplete method , there is absolutely no guarantee of finding a solution in a given limit of time .
remark that all unsatisfiable ( unforced ) instances below the threshold have been filtered out in order to make a fair comparison .
it appears that both complete and incomplete methods behave similarly . in figure
[ fig : tabu ] , one can see that the search effort grows exponentially with @xmath5 and that forced instances are as hard as unforced ones .
instances ) of solving instances in rb(@xmath34,@xmath72,@xmath73,@xmath74,@xmath8 ) using a tabu search , width=246 ]
as a related work , we can mention the recent progress on generating hard satisfiable sat instances .
@xcite have proposed to build random satisfiable 3-sat instances on the basis of a spin - glass model from statistical physics .
another approach , quite easy to implement , has also been proposed by @xcite : any 3-sat instance is forced to be satisfiable by forbidding the clauses violated by both an assignment and its complement .
finally , let us mention @xcite which propose to build random instances with a specific structure , namely , instances of the quasigroup with holes ( qwh ) problem .
the hardest instances belong to a new type of phase transition , defined from the number of holes , and coincide with the size of the backbone .
in this paper , we have shown , both theoretically and practically , that the models rb ( and rd ) can be used to produce , very easily , hard random instances .
more importantly , the same result holds for instances that are forced to be satisfiable . to perform our experimentation
, we have used some of the most efficient complete and incomplete csp solvers .
we have also encoded some forced binary csp instances of class rb(@xmath34,@xmath5,@xmath73,@xmath101,@xmath102 ) with @xmath5 ranging from @xmath103 to @xmath104 into sat ones ( using the direct encoding method ) and submitted them to the sat competition 2004 . about @xmath28 of the competing solvers
have succeeded in solving the sat instances corresponding to @xmath105 ( @xmath106 and @xmath107 ) whereas only one solver has been successful for @xmath108 ( @xmath109 and @xmath110 ) .
although there are some other ways to generate hard satisfiable instances , e.g. qwh @xcite or 2-hidden @xcite instances , we think that the simple and natural method presented in this paper , based on models with exact phase transitions and many hard instances , should be well worth further investigation .
the first author is partially supported by nsfc grant 60403003 and fanedd grant 200241 .
other authors have been supported by the cnrs , the `` programme cocoa de la rgion nord / pas - de - calais '' and by the `` iut de lens '' . | in this paper , we try to further demonstrate that the models of random csp instances proposed by @xcite are of theoretical and practical interest . indeed , these models , called rb and rd , present several nice features .
first , it is quite easy to generate random instances of any arity since no particular structure has to be integrated , or property enforced , in such instances .
then , the existence of an asymptotic phase transition can be guaranteed while applying a limited restriction on domain size and on constraint tightness . in that case , a threshold point can be precisely located and all instances have the guarantee to be hard at the threshold , i.e. , to have an exponential tree - resolution complexity .
next , a formal analysis shows that it is possible to generate forced satisfiable instances whose hardness is similar to unforced satisfiable ones .
this analysis is supported by some representative results taken from an intensive experimentation that we have carried out , using complete and incomplete search methods . |
a number of modified gravity theories have been proposed so far as alternatives to dark energy @xcite .
such theories are typically described by scalar - tensor gravity , in which a scalar propagating degree of freedom is introduced in addition to the gravitational wave ones .
a single scalar and two tensor modes imply that equations of motion for the scalar field and the metric are of second order . the horndeski theory of gravity @xcite
, the most general scalar - tensor theory with second - order field equations , is therefore quite useful for a comprehensive study of modified gravity .
recently , it has been noticed that the horndeski theory can be generalized without introducing any degrees of freedom other than the single scalar and two tensor modes @xcite .
going beyond the horndeski class , one inevitably obtains higher order field equations since the horndeski lagrangian has been proven to be the most general one with second - order field equations . the trick is to introduce a preferred time slicing on which the necessary number of initial conditions remains the same as that of the horndeski theory .
the preferred slicing is given by the unitary gauge in which the scalar field @xmath1 is spatially homogeneous , and away from the unitary gauge the field equations are of higher order in general .
one can thus obtain a more general framework to study modified gravity based on scalar - tensor theories beyond horndeski .
aspects of scalar - tensor theories beyond horndeski have been investigated in refs . @xcite . for a further generalization of the work of @xcite , see refs .
@xcite . the scalar degree of freedom in modified gravity is supposed to accelerate the current cosmic expansion , while care must be taken because the scalar - mediated force would persist down to small scales where a deviation from general relativity ( gr ) is strongly constrained , _
e.g. _ , by the solar - system tests . a screening mechanism for the scalar force on small scales is therefore required to be incorporated in modified gravity .
the vainshtein mechanism @xcite is one such screening mechanism that operates in theories possessing second derivatives of the scalar in the lagrangian .
the galileons @xcite , dgp gravity @xcite , and massive gravity @xcite are known to exhibit vainshtein screening in the vicinity of a gravitational source @xcite .
the original horndeski lagrangian can be written equivalently in the form of the generalized galileon @xcite , and this galileon - like structure allows us to study the vainshtein mechanism in a generic manner .
several works have been carried out along this direction : the generic conditions for the vainshtein mechanism to work were derived and some observational implications were suggested in @xcite , and the stability of the screened solutions was discussed in @xcite .
see also refs .
@xcite for related discussions .
the purpose of this paper is to extend those previous works and to study how the scalar - tensor lagrangian beyond horndeski changes the behavior of small - scale gravitational fields .
this paper is organized as follows . in the next section
we review how the horndeski theory can be generalized while maintaining the number of propagating degrees of freedom @xcite . in sec .
iii we derive the cosmological background equations in the generalized horndeski theory . in sec .
iv , the effective lagrangian that governs small - scale gravity on the cosmological background is obtained taking into account all relevant nonlinear terms .
we then discuss the possible impact of the new terms arising from the lagrangian beyond horndeski : first , we investigate linear density perturbations in sec .
v , and next we study the vainshtein mechanism for spherical overdensities in sec .
we draw our conclusions in sec .
the horndeski theory , _ i.e. _ , the most general scalar - tensor theory with _ second - order _ field equations , is described by @xmath2.\ ] ] the action controls the dynamics of gravity and the scalar field , @xmath1 , and its explicit expression is given by @xmath3 where @xmath4 , \\
l_5&:=&g_5(\phi , x)g_{\mu\nu}\nabla^\mu\nabla^\nu\phi -\frac{1}{6}g_{5x}\left [ ( \box\phi)^3 - 3\box\phi(\nabla_\mu\nabla_\nu\phi)^2 + 2(\nabla_\mu\nabla_\nu\phi)^3\right].\end{aligned}\ ] ] and each @xmath5 is an arbitrary function of @xmath1 and @xmath6 . throughout the paper , subscripts @xmath7 and @xmath1
are understood as differentiation , _
e.g. _ , @xmath8 . in the unitary gauge , @xmath9 , we have @xmath10 with @xmath11 being the lapse function , and
hence a function of @xmath1 and @xmath7 can be regarded as that of @xmath12 and @xmath11 . here and
hereafter a dot ( @xmath13 ) indicates a derivative with respect to @xmath12 .
this leads to the following unitary gauge description of the horndeski theory : @xmath14 with @xmath15 where @xmath16 and @xmath17 are the extrinsic and intrinsic curvature tensors of the spatial hypersurfaces , respectively . the central idea of ref .
@xcite is that one can maintain the number of the propagating degrees of freedom even if @xmath18 and @xmath19 are taken to be arbitrary functions of @xmath12 and @xmath11 rather than the functions constrained as eq .
( [ h - relation ] ) .
this generalization does not change the essential structure of the equation of motion for @xmath11 : it does not give the constraint among the dynamical variables , giving rise to one scalar degree of freedom on top of the usual gravitational wave modes .
ungauging the unitary gauge description ( [ uni - l ] ) with arbitrary @xmath18 and @xmath19 , one obtains the generalized horndeski theory with the following new terms @xcite : @xmath20 \biggr\ } \notag\\ & = & -\frac{1}{2}f_4(\phi , x ) \epsilon^{\mu\nu\alpha\beta } \epsilon_{\mu'\nu'\alpha'\beta}\nabla^{\mu'}\phi\nabla_\mu\phi \nabla^{\nu'}\nabla_\nu\phi\nabla^{\alpha'}\nabla_\alpha\phi,\label{eq : tilde l4 } \\
\widetilde{l}_5&:= & f_5(\phi , x ) \biggl\ { ( \box\phi)^2\nabla^\mu\phi\nabla^\nu\phi\nabla_{\mu}\nabla_\nu\phi -2\box\phi\nabla_\mu\phi\nabla^{\mu}\nabla^{\nu } \phi\nabla_{\nu}\nabla_{\lambda}\phi\nabla^\lambda\phi \nonumber\\ & & \quad\quad\quad -(\nabla_\mu\nabla_\nu\phi)^2\nabla^\rho\phi\nabla^\lambda\phi\nabla_\rho\nabla_\lambda\phi + 2\nabla_\mu\phi\nabla^\mu\nabla^\nu\phi\nabla_{\nu}\nabla_\rho\phi \nabla^\rho\nabla^\lambda\phi\nabla_\lambda\phi \nonumber\\ & & \quad\quad\quad + \frac{2}{3}x\left [ ( \box\phi)^3 - 3\box\phi(\nabla_\mu\nabla_\nu\phi)^2 + 2(\nabla_\mu\nabla_\nu\phi)^3\right ] \biggr\ } \notag\\ & = & -\frac{1}{3}f_5(\phi , x)\epsilon^{\mu\nu\alpha\beta } \epsilon_{\mu'\nu'\alpha'\beta'}\nabla^{\mu'}\phi\nabla_\mu\phi \nabla^{\nu'}\nabla_\nu\phi\nabla^{\alpha'}\nabla_\alpha\phi\nabla^{\beta'}\nabla_\beta\phi,\label{eq : tilde l5}\end{aligned}\ ] ] where @xmath21 is the totally antisymmetric levi - civita tensor density of weight @xmath22 .
even though adding these terms to the lagrangian results in _ third _ derivatives in the field equations , the number of propagating degrees of freedom remains unchanged .
this fact has recently been verified by the detailed hamiltonian analysis @xcite . in this paper , we work in the generalized horndeski theory with the matter sector minimally coupled to gravity , the action of which is given by @xmath23+s_{\rm matt}[g_{\mu\nu},\psi , a_{\mu},\cdots ] = \int{{\rm d}}t{{\rm d}}^3\mathbf{x}\sqrt{-g}\left ( l_{\rm h}+\widetilde{l}_4+\widetilde{l}_5\right ) + s_{\rm matt}. \label{eq : action}\end{aligned}\ ] ] the matter sector is composed of the standard model particles and cold dark matter , which form structures such as galaxies , galaxy clusters , and halos . before closing this section , it is worth noting that one can further generalize the above theory without introducing any extra degrees of freedom . indeed , following the spirit of the effective field theory approach it is possible to add various terms constructed from @xmath16 , @xmath17 , and @xmath24 having the coefficients that are functions of @xmath12 and @xmath11 @xcite
. this construction yields derivatives higher than three in the field equations in general , while not increasing the number of propagating degrees of freedom .
the resultant theory includes , for example , versions of hoava - lifshitz gravity @xcite as specific cases because now the combination of the form @xmath25 is allowed .
nevertheless , in this paper we focus on the theory introduced by gleyzes _
et al . _
@xcite as it can be regarded as the simplest extension of the horndeski theory .
let us begin with investigating the cosmological background equations .
the metric and the scalar field are spatially homogeneous and isotropic , @xmath26 and @xmath9 .
one can derive the field equations by varying the action with respect to @xmath11 , @xmath27 and @xmath1 , and then by setting @xmath28 .
the resultant equations are @xmath29 \nonumber\\ & & \quad + 8hx^2\left [ 2\left(h^2+\dot h\right)\dot\phi f_5 + 5h\ddot\phi f_5 + 2hx\ddot\phi f_{5x}+2hxf_{5\phi } \right]=-p,\label{ijb}\end{aligned}\ ] ] and @xmath30 with @xmath31 and @xmath32 we have included the energy density and pressure of usual matter in the right - hand sides of eqs .
( [ 00b ] ) and ( [ ijb ] ) . here , @xmath33 , @xmath34 , @xmath35 , and @xmath36 depend only on the horndeski functions @xmath5 , and
their explicit forms can be found in the literature @xcite .
( they are replicated in the appendix for the reader s convenience . )
one notices that @xmath33 is at most cubic in @xmath37 and no higher order terms are generated from @xmath38 and @xmath39 since the lagrangian beyond horndeski is at most cubic in @xmath16 ; the friedmann equation is still at most cubic in @xmath37 in the generalized horndeski theory . at sufficiently early times , we assume that the highest - order terms in @xmath37 are dominant in the friedmann equation and determine the expansion law of the universe . note that the coefficient of the cubic term is composed only of @xmath40 and @xmath41 .
therefore , in theories with @xmath42 and @xmath43 we have the early - time behavior as since terms with @xmath44 and @xmath45 can be integrated by parts and absorbed into those with @xmath46 and @xmath47 . ]
@xmath48 where @xmath49^{-1},\end{aligned}\ ] ] unless a tracker scaling solution is realized .
we thus see that in this case the scalar - field contribution to the cosmic expansion is screened at early times except for the time - dependence of @xmath50 . at late times
the other terms in the friedmann equation are assumed to become larger than the matter energy density to be responsible for the accelerated expansion .
for the purpose of investigating the vainshtein screening mechanism on small scales , we construct an effective theory for small fluctuations sourced by a matter overdensity @xmath51 , within which galaxies and associated objects are essentially frozen in their present - day configurations . to do so ,
let us consider quasi - static perturbations on a cosmological background .
working in the newtonian gauge , we write the ( dimensionful ) metric perturbations , the perturbed scalar field , and the nonrelativistic matter energy density as @xmath52 and @xmath53,\end{aligned}\ ] ] where @xmath54 is a mass scale which is assumed to be of order of the planck mass .
to study the quasi - static behavior of those perturbations on small scales , we expand the action ( [ eq : action ] ) in terms of @xmath55 and @xmath56 under the following rules .
we employ the quasi - static approximation , so that @xmath57 in the action , where @xmath58 stands for any of @xmath59 , @xmath60 , and @xmath56 , and @xmath61 denotes a spatial derivative.$ ] , where @xmath62 is the compton wavelength of the perturbations , the mass term is subleading and in particular screening due to @xmath56 s mass , _
i.e. _ , the chameleon mechanism @xcite can not take place .
the size of the overdensity region one considers must then be much smaller than the compton wavelength @xmath62 . ]
the matter overdensity @xmath51 is assumed to be of @xmath63 .
expanding the action gives rise to terms with second derivatives acting on @xmath58 .
in particular , we will have many nonlinear terms of the form @xmath64 or equivalently @xmath65 , and also @xmath66 ( @xmath67 ) .
we keep all such terms because they can be larger than the quadratic terms below a certain scale .
such nonlinear derivative interactions are crucial for realizing the vainshtein mechanism
. however , we ignore for example @xmath68 since it has less derivatives than @xmath69 and hence is smaller . by doing the expansion following the above rule
, we obtain the effective action of the form @xmath70 where @xmath71 is the interactions included in the horndeski class and @xmath72 is those beyond horndeski .
let us first look at the effective lagrangian derived from the horndeski part : @xmath73 where the quadratic and nonlinear terms are given by @xmath74-\pi_{ij } \right),\label{h - nl}\end{aligned}\ ] ] respectively . here , @xmath75 is the ( spatial ) galileon lagrangians and @xmath76 are the corresponding equations of motion derived from @xmath77 : @xmath78 , \\
{ \cal l}_4^{{\rm gal}}&=&-\frac{1}{2}(\nabla\pi)^2{\cal e}^{{\rm gal}}_3 , \\ { \cal e}_3^{{\rm gal}}&=&[\pi]^2-[\pi^2 ] , \\ { \cal e}_4^{{\rm gal}}&=&[\pi]^3 - 3[\pi][\pi^2]+2[\pi^3],\end{aligned}\ ] ] where we have introduced the convenient notation @xmath79=\delta^{ij}\pi^n_{ij}.\end{aligned}\ ] ] the quintic galileon lagrangian , @xmath80 , does not appear in the effective action because it is a total divergence in three dimensions .
( [ h-2 ] ) and ( [ h - nl ] ) we have defined the time - dependent dimensionless coefficients , @xmath81 , which can be written in terms of @xmath82 , @xmath83 , and @xmath84 .
their concrete form can be deduced from ref .
@xcite . for the purpose of the present paper
it is sufficient to write only some of the coefficients explicitly in terms of @xmath85 as summarized in appendix a. we assume that these coefficients are @xmath86 unless they vanish , based on the underlying assumption that @xmath85 scale as @xmath87 where @xmath88 is another mass scale . to explain the cosmic acceleration due to the interaction of the scalar field , it is reasonable to assume that @xmath88 is related to the present value of the hubble rate , @xmath89 , as @xmath90 .
then it is easy to see that the scaling is consistent with the cosmological background equations .
a concrete realization of the above scaling and self - accelerating cosmology with @xmath91 can be found for example in massive gravity @xcite .
the effective lagrangian ( [ eq : lag_h_eff ] ) is the cosmological generalization of ref .
@xcite in the newtonian gauge , and the corresponding equations of motion have already been derived in ref .
@xcite .
the two lagrangians beyond horndeski , @xmath38 and @xmath39 , generate the following new terms , @xmath92\pi_{ij}-\pi^2_{ij}\right),\label{beyond - nl}\end{aligned}\ ] ] where we have introduced the dimensionless parameters of @xmath86 ( unless they vanish ) as @xmath93 with @xmath94 being a mass scale related to @xmath88 through @xmath95 .
here we have assumed the scaling @xmath96 this scaling is consistent with the cosmological background equations .
note that the two lagrangians also generate the same horndeski terms as appearing in eqs .
( [ h-2 ] ) and ( [ h - nl ] ) , which can be absorbed in the redefinition of the corresponding coefficients .
we thus move all the horndeski - type interactions arising from @xmath38 and @xmath39 into @xmath97 , and define @xmath98 solely in terms of the new interactions as @xmath99 . from eqs .
( [ beyond-2 ] ) , and ( [ beyond - nl ] ) one immediately notices that field equations beyond horndeski involve third derivatives of the form @xmath100 and @xmath101 .
the time - derivative terms here can not be ignored even though one employs quasi - static approximation .
for example , @xmath102 must be retained in the equations of motion which will be of the same order as @xmath103 .
one sees that only in the case where @xmath104 the new interactions ( [ beyond-2 ] ) and ( [ beyond - nl ] ) participate in determining small - scale gravitational fields .
hence , taking the limit @xmath105 , @xmath106 , only the horndeski terms ( [ h-2 ] ) and ( [ h - nl ] ) survive , and one finds from the concrete expression for the coefficients that @xmath107 , @xmath108 , @xmath109 , and @xmath110 .
the form of the resultant effective lagrangian exactly coincides with the newtonian gauge expression of the effective theory of the vainshtein mechanism from the horndeski theory derived in ref .
@xcite with time - derivative terms being dropped . on the cosmological background , however , the above degeneracy of the coefficients
is disentangled in general .
note that , even in the case of a minkowski background , one can consider a solution with @xmath111 const @xmath112 ( _ i.e. , _
@xmath113 ) in a certain subclass of the general scalar - tensor theories , and the degeneracy of the coefficients is disentangled around such a background as well due to the preferred time direction specified by @xmath114 .
on such a background @xmath115 does not vanish . to summarize , the total effective lagrangian is given by @xmath116 where the form of each lagrangian in the right - hand side is found in eqs .
( [ h-2 ] ) , ( [ h - nl ] ) , ( [ beyond-2 ] ) , and ( [ beyond - nl ] ) .
( now the coefficients in @xmath117 and @xmath118 are composed not only of @xmath85 but also of @xmath119 due to the above - mentioned redefinition of the coefficients . )
to explore the impact of the term @xmath120 , let us focus on the evolution of the density perturbation @xmath51 in the linear regime in this section , deferring the analysis of the nonlinear regime until sec . vi .
since the matter sector is assumed to be _ minimally coupled _ to gravity , the evolution equations for @xmath51 is the same as the conventional one , @xmath121 this equation is a consequence of the conservation equations for the matter energy - momentum tensor , and therefore can be derived without using the field equations for the metric and the scalar field . to close the evolution equation one needs to relate @xmath59 to @xmath51 , and
this is where the effect of modified gravity comes in . in deriving the effective action on small scales in the previous section
, we made the quasi - static approximation @xmath122 in the action .
working within the horndeski theory , this approximation results in three algebraic equations for @xmath123 , @xmath124 , and @xmath125 at linear order @xcite .
we can thus write those three variables in terms of @xmath51 , and the evolution equation for @xmath51 is now closed .
note that @xmath126 , and hence the time derivatives of @xmath51 are retained in eq .
( [ ev - density ] ) .
since we have a single second - order differential equation for @xmath51 , we need two initial conditions to determine the quasi - static evolution of @xmath51 in the horndeski theory .
going beyond the horndeski theory , @xmath120 gives rise to terms of @xmath127 in the field equations .
since the interactions between the metric and the scalar field are rather complicated , it is difficult to investigate the equations of motion in the original jordan frame .
a simple way of dealing with this is to define the einstein - frame variables as @xmath128 where @xmath129,\\ b_2&:=&\frac{\xi_t}{{\cal g } } , \end{aligned}\ ] ] and we have assumed that @xmath130 and @xmath131 . in terms of @xmath132 and @xmath133 , the quadratic part of the lagrangian in the einstein frame is written as @xmath134 and we redefined the coefficient of the @xmath135 term : @xmath136 we can thus remove @xmath120 as well as coupling between the metric perturbations and @xmath56 to go to the einstein frame . in this frame , however , @xmath56 couples to matter .
in particular , @xmath120 yields the `` disformal '' piece @xmath137 . working in the einstein - frame lagrangian
, the kinetic term for small fluctuations has the right sign provided that @xmath138 .
the field equations derived from the lagrangian are given by @xmath139 where we have used the fact that @xmath140 for nonrelativistic matter .
substituting the definitions into these equations and using the evolution equation , we see that @xmath123 in the original jordan frame can be written in the form @xmath141 where @xmath142 \right\ } \,,\\ & & \alpha:=\left(1+\frac{4b_2 ^ 2\rho}{{\widetilde{m}_{\rm pl}}^2{\cal m}^2\hat\eta}\right)^{1/2}a.\label{modd2}\end{aligned}\ ] ] one can easily see that the cosmological poisson equation is modified due to the additional friction term from the scalar field , as well as the time - dependent effective gravitational coupling . within the horndeski theory ( @xmath143 ) , it is clear that @xmath144 reduces to @xmath27 . in this case
, @xmath145 simply represents the effective gravitational coupling for linear density perturbations , the explicit form of which can be found in ref .
@xcite . combining eqs .
( [ ev - density ] ) and ( [ ddphi ] ) , a closed second - order differential equation for @xmath51 is obtained .
going beyond horndeski , it turns out from eq .
( [ ddphi ] ) that the evolution equation for @xmath51 takes the form @xmath146 equations and imply that , in addition to the usual change of the effective gravitational coupling , @xmath120 effectively modifies the cosmic expansion rate felt by the density perturbation .
note that as long as the stability condition for @xmath56 is imposed , the singular point of this equation , @xmath147 , is obviously avoided for @xmath148 . the evolution equation for the matter overdensity with the same structure as eq .
( [ modd1 ] ) has also been derived in the context of a disformally coupled scalar field @xcite .
in the previous section we have investigated the impact of the new interaction beyond horndeski on the quasi - static evolution of linear density perturbations .
now we are going to study the nonlinear terms beyond horndeski and their possible impact on the vainshtein mechanism , assuming spherical symmetry for simplicity .
we focus on the subclass of the theory in which @xmath149 .
this restriction is not only for simplicity , but also for the following reasons : within the horndeski theory it is pointed out that spherically symmetric vainshtein solutions are unstable against perturbations if @xmath150 @xcite , and the metric potentials do not show the correct newtonian behavior @xmath151 on small scales on the cosmological background if @xmath152 @xcite . in terms of the functions in the original lagrangian of the theory ,
the restriction @xmath149 amounts to switching off @xmath40 and @xmath41 .
it follows then that @xmath153 .
note also that in @xmath154 theories the friedmann equation contains no cubic term in @xmath37 , and hence conventional cosmology is recovered ( except that the effective gravitational `` constant '' @xmath50 is time - dependent in general ) for @xmath155 .
varying @xmath156 with respect to @xmath60 , @xmath59 , and @xmath56 , one obtains the field equations valid on small scales , which can be integrated once immediately .
the relevant equations are summarized as follows : @xmath157 where we have defined the dimensionless quantities as @xmath158 with @xmath159 being the enclosed mass . here
@xmath160 is the comoving radial coordinate , and a prime ( @xmath161 ) indicates a partial derivative with respect to @xmath160 . upon integration
all the integration constants are set to @xmath162 because @xmath163 at infinity . unlike theories within horndeski ,
@xmath164 theories generate second derivatives in eqs .
( [ eq1 ] ) and ( [ eq3 ] ) , since the field equations away from the unitary gauge are of third order in general . unlike the analysis of linear density perturbations , we will not consider a disformal transformation of the metric because the metric would no longer take the newtonian gauge form after such a transformation at nonlinear order . in the region
far from the source , we have @xmath165 and linearization is justified to get the solution @xmath166 . in this section ,
we consider the region in the vicinity of the source satisfying @xmath167 , where nonlinearity , and hence possible new effects from @xmath115 , are expected to play a crucial role .
let us now identify the appropriate solution for @xmath167 . using eqs .
( [ eq1 ] ) and ( [ eq2 ] ) , one can eliminate @xmath168 and @xmath169 from eq .
( [ eq3 ] ) .
interestingly , in doing so @xmath170 and @xmath171 drop out from eq .
( [ eq3 ] ) , and we have the equation of the form @xmath172 + 2\left[\kappa_1 + \left({\cal f}\alpha_1-{\cal g}\alpha_2 + 3{\cal g}\alpha_\ast\right)a + { \cal g}\alpha_\ast ra ' \right ] x+\kappa_2 x^2- \xi\ , x^3=0 , \label{eq : x eq}\end{aligned}\ ] ] where @xmath173 we look for solutions with @xmath174 for @xmath167 , because for such solutions @xmath175 is suppressed due to nonlinearity so that we may expect that gr is reproduced inside the vainshtein radius . we find the following approximate solution for @xmath167 satisfying this requirement : @xmath176}{\xi}+{\cal o}(a^{1/2}),\label{sol1}\end{aligned}\ ] ] where we have assumed that @xmath177 . substituting eq .
( [ sol1 ] ) into eqs .
( [ eq1 ] ) and ( [ eq2 ] ) , we obtain @xmath178 using the concrete expression of the coefficients in terms of @xmath179 and @xmath180 , eqs .
( [ ysol1 ] ) and ( [ zsol1 ] ) can be recast in a more suggestive form as @xmath181 where it turns out that @xmath182 let us consider a density profile for which @xmath183 outside a certain radius @xmath184 . following refs .
, it will be appropriate to define the vainshtein radius as @xmath185^{1/3}.\end{aligned}\ ] ] then , outside the overdensity ( @xmath186 ) , we have @xmath187 since @xmath188 outside the source , the second terms in eqs .
( [ ysol2 ] ) and ( [ zsol2 ] ) vanish for @xmath189 , leading to @xmath190 , _
i.e. , _ @xmath191 this implies that gr is reproduced inside the vainshtein radius and outside the overdensity , provided that @xmath192 varies in time sufficiently slowly .
note in passing that @xmath193 as the dimensionless coefficients have been assumed to be of @xmath86 .
we thus have looked at the region outside the overdensity and found that the @xmath194 term does not change the qualitative aspect of gravity near the source : in the absence of @xmath195 ( and @xmath196 ) the vainshtein mechanism operates to reproduce standard gravity , @xmath197 , but the effective gravitational coupling is time - dependent in general and given by @xmath50 @xcite .
the impact of the new terms beyond horndeski becomes manifest only inside the overdensity region , where the second terms in eqs .
( [ ysol2 ] ) and ( [ zsol2 ] ) no longer vanish . when this is the case , the gravitational attraction is determined not only from the enclosed mass ( @xmath198 ) but also from the local energy density and its derivative ( @xmath199 , @xmath200 ) .
it is clear from eqs .
( [ ysol2 ] ) and ( [ zsol2 ] ) that @xmath201 for @xmath202 , and hence the two metric potentials do not coincide : @xmath203 which implies that _ the vainshtein screening mechanism fails to operate inside the overdensity region in the presence of the interaction beyond horndeski . _ to estimate the difference between the two potentials ,
let us consider for simplicity the case with @xmath204const . in this case
it is easy to see that @xmath205 and hence @xmath206 in general .
breaking of the screening mechanism inside the overdensity region could for example affect the motion of a galaxy inside a cluster and modify structure of stars , though one should take into account the effect of pressure in the stellar interior .
astrophysical implications of this novel modification of gravity are beyond the scope of the present paper and will be discussed in more detail elsewhere . , @xmath168 , @xmath169 .
the profile @xmath207 is such that @xmath208 const for @xmath209 ( i.e. , @xmath210 corresponds to the surface of a constant density star ) , as shown by the thin solid line .
we assume that the system is static , and the coefficients are given by @xmath211 , @xmath212 , @xmath213 , @xmath214 , @xmath215 , @xmath216 , @xmath217 , @xmath218 , @xmath219 , and @xmath220 .
the vainshtein mechanism works well for @xmath221 , giving @xmath222 .
however , for @xmath223 we find @xmath224 , while for @xmath225 we have @xmath226 , and hence standard gravity is not recovered there . , height=340 ] a numerical example is presented in fig .
[ fig : fig.eps ] , which agrees well with our analytic estimate . for completeness ,
let us make a brief comment on another solution of eq .
( [ eq : x eq ] ) for @xmath167 .
the solution is of @xmath86 and is given by @xmath227 where we have assumed that @xmath177 and @xmath228 .
it can be seen immediately from eqs .
( [ eq1 ] ) and ( [ eq2 ] ) that for this @xmath229 , @xmath230 therefore , @xmath231 in general even in the vainshtein radius .
, corresponds to @xmath232 , where @xmath233 is the propagation speed of gravitational waves . in the presence of the terms beyond horndeski ,
the propagation speed is given in general by @xcite @xmath234,\end{aligned}\ ] ] and hence the condition @xmath107 does not correlate with the speed of gravitational waves . ]
this property is independent of whether the region inside or outside the overdensity is considered .
in this paper , we have derived an effective theory on small scales from scalar - tensor theories beyond horndeski @xcite . in doing so
, we have made a quasi - static approximation in the newtonian gauge , while keeping all the nonlinear terms which could be relevant on small scales .
we have seen that the new interaction terms beyond horndeski give rise to third derivatives in the field equations .
the impact of those new terms has been investigated in the case of @xmath235 theories .
we have found that the linear growth of matter density perturbations is modified not only through the time - dependent effective gravitational coupling but also through the additional friction , which is absent in the horndeski theory .
this is due to the disformal coupling of the scalar field to matter in the einstein frame .
we have then investigated the nonlinear effect of the scalar - field fluctuations that can screen the fifth force inside the vainshtein radius .
one of the solutions outside and near the source has been shown to reproduce the standard behavior , @xmath236 , though the effective gravitational coupling , @xmath192 , is time - dependent in general on a cosmological background , as in the horndeski theory . in particular , @xmath192 coincides with the gravitational coupling in the friedmann equation even in the presence of the @xmath194 term .
however , the new interactions beyond horndeski crucially change the behavior of the gravitational potentials inside the matter overdensity .
the gravitational attraction depends not only on the enclosed mass but also on the local matter energy density , and @xmath59 and @xmath60 no longer coincide , implying that gr is not recovered inside the source .
we have made the quasi - static approximation for scalar perturbations throughout the paper .
this hindered us from investigating the stability of the spherical solutions in sec .
. it would be interesting to study how the new terms beyond horndeski change the stability _
e.g. _ of stars .
this work was supported in part by jsps grant - in - aid for young scientists ( b ) no .
24740161 ( t.k . ) , the jsps research fellowship for young scientists nos .
269337 ( y.w . ) and 259800 ( d.y . ) , and the munich institute for astro- and particle physics ( miapp ) of the dfg cluster of excellence `` origin and structure of the universe '' ( y.w . ) .
the cosmological background evolution in the horndeski theory is given in terms of the gravitational analogue of the energy density @xmath33 , the isotropic pressure @xmath34 , and the current @xmath35 of the scalar field @xmath1 @xcite : @xmath237 and @xmath238 where @xmath239g_{5\phi } + 4hx\dot\phi g_{5\phi\phi}.\end{aligned}\ ] ] the equation of motion for the scalar field is given by @xmath240 where @xmath241 we basically follow the notations in refs .
@xcite rather than those in ref .
@xcite . here
we summarize the relations between the coefficients in the effective action and those of the equations of motion derived in ref .
@xcite . in the absence of the terms beyond horndeski , the coefficients in eqs .
( [ h-2 ] ) and ( [ h - nl ] ) are given by @xmath242 where @xmath243 , @xmath244 , ... in the right - hand sides are found in appendix a of ref .
the lagrangians @xmath38 and @xmath39 generate the horndeski terms as well as the new terms ( [ beyond-2 ] ) and ( [ beyond - nl ] ) . due to this ,
the above coefficients are redefined as @xmath245 we investigate the theories with @xmath154 in the main text . in this case
it can be seen that @xmath246 .
one may also notice that the following relation holds : @xmath247 c. deffayet , x. gao , d. a. steer and g. zahariade , phys .
d * 84 * , 064039 ( 2011 ) [ arxiv:1103.3260 [ hep - th ] ] .
t. kobayashi , m. yamaguchi and j. yokoyama , prog .
phys . *
126 * , 511 ( 2011 ) [ arxiv:1105.5723 [ hep - th ] ] . | the horndeski theory of gravity is known as the most general scalar - tensor theory with second - order field equations .
recently , it was demonstrated by gleyzes _
et al .
_ that the horndeski theory can further be generalized in such a way that although field equations are of third order , the number of propagating degrees of freedom remains the same .
we study small - scale gravity in the generalized horndeski theory , focusing in particular on an impact of the new derivative interaction beyond horndeski on the vainshtein screening mechanism . in the absence of the quintic galileon term and its generalization ,
we show that the new interaction does not change the qualitative behavior of gravity outside and near the source : the two metric potentials coincide , @xmath0 , while the gravitational coupling is given by the cosmological one and hence is time - dependent in general .
we find , however , that the gravitational field inside the source shows a novel behavior due to the interaction beyond horndeski : the gravitational attraction is not determined solely from the enclosed mass and two potentials do not coincide , indicating breaking of the screening mechanism . |
among the various issues that remain open in the field of accretion onto black hole binaries ( bhbs ) as well as super - massive black holes is the way the gas accretes at very low eddington ratios , in the so called _ quiescent _ regime ( where , for the binaries , the boundary between quiescence and a more active regime can be set around @xmath3 erg , corresponding to a few @xmath4for a 10 bh ; the reader is referred to mcclintock & remillard @xcite for a review of x - ray states ) . while there is little doubt that the thin disc model @xcite captures the basic physical properties of bhbs in the thermal dominant state , the accretion mode responsible for powering quiescent bhs is still a matter of debate .
observations of highly sub - eddington systems , most notably the galactic center super - massive bh , paved the way to radiatively inefficient accretion flow models ( riafs ) . by reviewing the vast literature on the subject , one immediately comes to the conclusion that the most widely accepted / adopted model for reproducing the spectral energy distribution of quiescent bhs is the ` advection - dominated accretion flow ' solution ( adaf @xcite@xcite ) ; in this low - density , two temperature inflow , a significant fraction of the viscously dissipated energy remains locked up in the ions as heat , and is advected inward . the adaf model can successfully account for the overall shape of the uv - optical - x - ray spectra of quiescent bhbs ( see mcclintock 2003 @xcite for an application to the high quality data of xte j1118 + 480 ) .
nevertheless , alternative suggestions are worth being considered . for instance , convection , rather than advection , could be responsible for keeping the plasma circulating in eddies along the flow by transporting angular momentum inward , yielding zero net accretion rate ( cdaf ; @xcite ) .
a different solution is that elaborated by blandford & begelman @xcite , where the excess energy and angular momentum is lost to an outflow at all radii ; the final accretion rate into the hole may be only a tiny fraction of the mass supply at large radii ( adios : ` adiabatic inflow - outflow solution ' ) .
radiatively inefficient outflows may also be responsible for dissipating the bulk of the liberated accretion power ( e.g. @xcite,@xcite,@xcite ) .
roughly speaking , the question remains open whether quiescent bhbs are dim because of a highly reduced radiative efficiency , or a highly reduced inner accretion rate .
recent chandra observations of nearby elliptical galaxies @xcite led to the discovery of a tight correlation between the bondi accretion rate , @xmath5 ( as inferred from the measured temperature , bh mass and density profile ) and the power emerging from these systems in the form of relativistic jets ( estimated by the work they exert on the observed x - ray cavities ) . the jets power is found to be comparable to @xmath6 , implying that a significant fraction of the matter entering the accretion radius flows down to regions close to the black holes , where the jets are thought to be formed .
in the context of x - ray binaries , as well as super - massive black holes , the term ` jet ' is typically used as a synonymous for a relativistic outflow of plasma and implies a high degree of collimation ( see fender 2006 @xcite for a comprehensive review on x - ray binary jets ) . as a matter of fact
, high spatial resolution radio observations of bhbs in the hard state @xcite have imaged highly collimated structures in two systems only : cyg x-1 @xcite and grs 1915 + 105 @xcite are both resolved into elongated radio sources on milliarcsec scales that is tens of a.u . implying collimation angles smaller than a few degrees on much larger scales than the orbital separation .
both systems display relatively high x - ray ( and radio ) luminosities , with grs 1915 + 105 being persistently close to the eddington limit ( e.g. @xcite ) , and cyg x-1 having a bolometric x - ray luminosity around 2 per cent of @xmath7 @xcite .
this , however , should not be taken as evidence against collimated jets at lower luminosities : because of sensitivity limitations on current high resolution radio arrays , resolving a radio jet at microjy level ( if any ) simply constitutes an observational challenge .
in addition , at such low levels , the radio flux could be easily contaminated by synchrotron emission from the donor star .
the presence of a collimated outflow can also be inferred by its long - term action on the local interstellar medium , as in the case of the hard state bhbs 1e1740.7@xmath82942 and grs 1758@xmath8258 , both associated with arcmin - scale radio lobes @xcite , @xcite .
further indications come from the stability in the orientation of the electric vector in the radio polarization maps of gx 339@xmath84 over a two year period @xcite .
this constant position angle , being the same as the sky position angle of the large - scale , optically thin radio jet powered by gx 339@xmath84 after its 2002 outburst @xcite , clearly indicates a favoured ejection axis in the system .
on the other hand , failure to image a collimated structure in the hard state of xte j1118 + 480 down to a synthesized beam of [email protected] mas@xmath10 at 8.4 ghz @xcite can challenge the collimated jet interpretation , even though xte j1118 + 408 was observed at roughly one order of magnitude lower luminosity with respect to e.g. cyg x-1 .
if the jet size scaled as the radiated power , one could expect the jet of xte j1118 + 408 to be roughly ten times smaller than that of cyg x-1 ( which is 2@xmath96 mas@xmath10 at 9 ghz , at about the same distance ) , still point - like in the vlba maps presented by mirabel 2001 @xcite .
garcia 2003 @xcite have pointed out that long period ( @xmath11 day ) bhbs undergoing outbursts tend to be associated with spatially resolved optically thin radio ejections , while short period systems would be associated with unresolved , and hence physically smaller , radio ejections .
if a common production mechanism is at work in optically thick and optically thin bhb jets @xcite , then the above arguments should apply to steady optically thick jets as well , providing an alternative explanation to the unresolved radio emission of xte j1118 + 480 , with its 4 hour orbital period , the shortest known for a bhb
. it is worth mentioning that , by analogy , a long period system , like for instance v404 cyg , might be expected to have a more extended optically thick jet ( approved high spatial resolution observations of this system , with the high sensitivity array , will hopefully answer this question ) .
in an attempt to assess the relation between accretion and jet production in hard state bhbs , corbel 2003 , and gallo , fender & pooley 2003 ( @xcite,@xcite ) have established the existence of a tight correlation between the x - ray and the radio luminosity ( @xmath12 and @xmath13 ) , of the form @xmath14 . in those works ,
10 hard state systems with nearly simultaneous radio / x - ray observations were considered ; the correlation was found to hold over more than 3 orders of magnitude in @xmath15 , up to a few per cent of @xmath16 , above which the sources enter the thermal dominant state , and the core radio emission drops below detectable levels .
probably the most notable implication of the non - linear scaling was the predicted existence of a critical x - ray luminosity below which a significant fraction of the liberated accretion power is channelled into a radiatively inefficient outflow , rather than being dissipated locally by the inflow of gas and emitted in the form of x - rays ( this does not necessarily imply the x - ray spectrum is dominated by non - thermal emission from the jet , as most of the jet power may be stored as kinetic energy @xcite ) .
due to its low x - ray luminosity ( @xmath17 @xcite ) and relative proximity , the bhb in represented the most suitable known system to probe the radio / x - ray correlation beyond the hard state . indeed , deep vla observations , performed in 2005 august , resulted in the first radio detection of a quiescent bhb emitting at such low x - ray luminosities ; the level of radio emission 51 @xmath18jy at 8.5 ghz is the lowest ever measured in an x - ray binary . at a distance of 1.2 kpc , this corresponds to a radio luminosity @xmath19 erg sec@xmath2 . by analogy with higher luminosity systems ,
partially self - absorbed synchrotron emission from a relativistic outflow appears as the most likely interpretation .
free - free wind emission is ruled out on the basis that far too high mass loss rates would be required , either from the companion star or the accretion disc , to produce observable emission at radio wavelengths , while gyrosynchrotron radiation from the corona of the companion star is likely to contribute to less than 5 per cent to the measured flux density .
a062000 was observed simultaneously in the x - ray band with , its 0.3 - 8 kev spectrum well fitted by an absorbed power law with photon index @xmath20 and hydrogen equivalent column density consistent with the optical value .
the corresponding 210 kev luminosity was about @xmath21 erg sec@xmath2 , a factor of two higher than in a previous observation , in february 2000 .
the simultaneous observations of allowed us to test and extend the radio / x - ray correlation for bhbs by 3 orders of magnitude in @xmath22 .
the measured radio / x - ray fluxes seem to confirm the existence of a non - linear scaling between the radio and x - ray luminosity in this systems ; with the addition of the point @xmath23 provides a good fit to the data for @xmath22 spanning between @xmath24 and @xmath25 ( see figure [ fig:1 ] ) . the fitted slope , albeit consistent with the previously reported value of @xmath26 , is admittedly affected by the uncertainties in the distance to gx3394 , for which the correlation extends over 3 orders of magnitude in @xmath22 and holds over different epochs .
even though the a062000 data appear to follow the hard state radio / x - ray correlation , since 2003 , when the compilation of quasi - simultaneous radio / x - ray observations of hard state bhbs was presented @xcite , many outliers have been found . to mention a few ( those that came or were brought to the attention of this author at least ) : xte j1720318 @xcite,@xcite , swift j1753.50127 @xcite , igr j174972821 @xcite and xte j1650500 @xcite , while in the hard state , all appear to lie significantly below the best - fitting correlation ( figure [ fig:1 ] ) .
while a number of plausible ad - hoc ` reasons ' can be adduced on a source - by - source basis , it simply seems more reasonable to conclude that the radio / x - ray correlation may not be universal , or , at least , that there is no universal normalization .
this casts doubts on the possibility of relying on the best - fitting relation for estimating other quantities , such as distance or bh mass .
a global re - assessment of the radio / x - ray correlation will be presented in a forthcoming paper .
while adaf models predict the existence of bipolar outflows emanating from the surface layers of the equatorial inflow @xcite , generally they do not address the importance of such outflows with respect to the overall accretion process in terms of energetics .
ideally , one would like to be able to compare the total power carried away in the form of outflows / jets , with the total accretion power budget available to the black hole . as mentioned in the first section ,
this has been done for a handful of nearby elliptical galaxies @xcite ; there , the mechanical jets power is found to be comparable to @xmath27 .
this has been interpreted as evidence that a substantial fraction of the captured mass does reach the innermost regions of the flow before being ejected in the form of an outflow . +
a similar conclusion has been reached in the case of the stellar mass bh in a062000 @xcite . based on models for the optical / uv emission of the outer accretion disc in dwarf novae @xcite ,
corrected downward to account for the mass difference , mcclintock 1995 @xcite estimate @xmath28 yr@xmath2 for a062000 , where @xmath29 is a factor of the order unity , that can be up to a few .
this value is also comparable to the @xmath30 yr@xmath2 inferred from the measure of the total energy released during the 1975 outburst of adopting 58 year recurrence time based on plate archives which showed an outburst in 1917 .
this time - averaged value had been calculated based on a distance of 1 kpc for ( vs. a refined value of 1.2 kpc ) and could still be underestimated by a factor 2 or so , to allow for the possibility that an intermediate outburst was missed .
the putative luminosity associated with @xmath31 , if it was to reach the black hole with standard radiative efficiency , would be @xmath32 erg sec@xmath2 , five orders of magnitude than the observed x - ray ( or bolometric ) luminosity . in the above expression @xmath33
is the accretion efficiency , which depends only on the bh spin .
the various riafs provide different explanations for the much lower luminosities that are observed in terms of different ` sinks ' for the energy . in the adaf scenario
, it is assumed that all the @xmath31 is accreted onto the black hole ( @xmath34 ) while the radiated luminosity @xmath35 is much smaller as a result of a reduced radiative efficiency @xmath36 where @xmath37 is the critical rate above which the disc becomes radiatively efficient .
the index @xmath38 is typically close to unity , but its exact value may depend on the micro - physics of adaf and on how the bolometric luminosity is calculated . writing the bolometric luminosity of as @xmath39 erg sec@xmath2 , being @xmath40 a multiplicative factor , then @xmath41 . for @xmath42 , @xmath43 yr@xmath2 , which corresponds to an eddington - scaled critical accretion rate @xmath44 . with @xmath45 ,
as implied by the adaf spectral modelling , a self - consistent adaf solution is obtained for @xmath46 of a few times @xmath47 , as expected on theoretical grounds .
contrary to the adios case , for the adaf scenario to be self - consistent , the total kinetic power of the jet / outflow , @xmath48 , should be a negligible fraction of @xmath49 . in order to verify this ,
@xmath50 can be estimated making use of the normalization for the jet kinetic power vs. radio luminosity derived by @xcite .
obviously , if this number is high enough , the jet contribution to the flow energetics ( and dynamics ) can be negligible with respect to advective cooling . in @xcite the radio core emission of three well studied radio galaxies ( m87 , per a and cyg a )
was directly compared to the radio lobe emission , used a jet calorimeter .
they proposed that the jet kinetic power can be estimated from the core radio luminosity in the following way : @xmath51 erg sec@xmath2 where @xmath52 is the radio spectral index , and the parameter @xmath53 carries the ( quite large ) uncertainty on the radio galaxy calibration . for , assuming a flat radio spectral index @xmath54 , and for @xmath55 erg sec@xmath2 , there follows : @xmath56 erg sec@xmath2 , or @xmath57 .
this would mean that the jet / outflow carries a significant amount of the accretion energy budget away from the system .
i so , then any realistic accretion flow model for quiescence shall necessarily incorporate the effects of an outflow both in terms of energetics and dynamics , effectively ruling out a pure adaf solution . using the estimate for the kinetic power of the jet in quiescence
, we can calculate the total energy carried out by it in between outbursts , assuming again that the 58 years recurrence time is not overestimated .
we obtain @xmath58 erg , of the same order of the energy released during an outburst .
interestingly , @xcite calculated the outburst evolution for with a model accounting for evaporation of the cold outer disc ( but neglecting outflows ) , and concluded that only about one third of the mass accreted during quiescence needs to be stored in the disc for the subsequent outbursting episode . under the adios working - hypothesis , the mass flowing from the outer disc does not reach the inner region , but is lost in a outflow .
the accretion rate is now a function of radius : @xmath59 where we have assumed that mass loss sets in within the truncation radius @xmath60 .
proceeding as before , we can then estimate the truncation radius for @xmath61=@xmath62 , being @xmath63 the schwarzschild radius for a 10 bh .
for @xmath64 we obtain @xmath65 or @xmath66 cm , to be compared with the orbital separation of about @xmath67 cm @xcite .
assuming that the the jet / outflow is powered by the mass lost from the accretion flow , then its total kinetic power @xmath50 is given by @xmath68 i.e. in the adios framework , a dominant fraction of the total accretion power is channelled into the jet / outflow . in conclusion , by making use of the estimate of the outer accretion rate of a062000 in quiescence and of the jet radiative efficiency by @xcite
, it has been argued that the outflow kinetic power accounts for a sizable fraction of the accretion energy budget , and thus must be important with respect to the overall accretion dynamics of the system @xcite .
on the other hand , as noted in the case of elliptical galaxies studied with chandra @xcite , the large jets mechanical power ( compared to the available fuel supply ) suggests that , before being expelled and carrying away a substantial fraction of the dissipated accretion power , the accreted mass must flow deep into the hole s potential well , were high efficiencies can be reached .
[ fig:1 ] a062000 @xcite ( open circles ) .
while this low eddington ratio system appears to confirm the validity of a non - linear radio / x - ray correlation over more than 6 orders of magnitude in @xmath12 , an increasing numbers of hard state outliers is being found at higher luminosities ( filled triangles ) , whose radio flux is typically lower with respect to the best - fitting power - law ( e.g. xte j1720318 @xcite,@xcite and swift j1753.50127 @xcite , igr j174972821 @xcite , xte j1650 - 500 @xcite ) .
this challenges the validity of such relation , or at least the universality of its normalization.,title="fig : " ] i wish to thank the organizers for putting together , once again , such an interesting conference in such a spectacular location .
this work is supported by nasa through chandra postdoctoral fellowship award pf5 - 60037 , issued by the chandra x - ray observatory center , which is operated by the smithsonian astrophysical observatory for and on behalf of nasa under contract nas8 - 39073 . | the question whether quiescent black hole x - ray binaries are capable of powering relativistic outflows is addressed by means of simultaneous radio / x - ray observations of a nearby system steadily emitting x - rays below @xmath0 times the eddington luminosity .
the robust detection of a radio counterpart suggests that a synchrotron emitting outflow is being powered by this system , even though its degree of collimation remains unknown , and hard to investigate . with the inclusion of the a062000 data ,
the non linear radio / x - ray correlation for hard state black hole x - ray binaries appears to hold down to very low quiescent luminosities .
however , an increasing number of outliers is being found at higher luminosities , questioning the universality of such correlation , or at least its normalization .
2@xmath1 1s@xmath2 2@xmath1 address = physics department , university of california , santa barbara , ca 93106 - 9530 , usa , altaddress = chandra fellow |
isolated pulsars are spinning neutron stars whose observed spin rates gradually decrease with time .
the age @xmath2 of a pulsar is usually assumed to be equal to the timing ( or characteristic ) age @xmath3 derived by assuming pure magnetic dipole spin - down _ in vacuo _ , and the age is then given by @xmath4 ( e.g. @xcite ) , where @xmath5 and @xmath6 are the angular spin frequency and angular frequency derivative for a pulsar spin period @xmath7 and period derivative @xmath8 .
under this same assumption of mdr spin - down , the magnetic field strength of the pulsar is given by the formula ( @xcite ) @xmath9 which is often assumed to be equal to the true field strength of the isolated pulsar ( e.g. @xcite ) .
psr b175724 is a @xmath10 s radio pulsar which appears to be associated with the supernova remnant g5.41.2 ( @xcite ) .
the pulsar is surrounded by a compact radio nebula having a cometary morphology with a tail extending back into the supernova remnant ( @xcite ) strongly suggesting that the pulsar was formed in the supernova which produced g5.41.2 ( @xcite ) .
given the temporal parameters of psr b175724 ( @xmath11 s and @xmath12 s s@xmath13 ; @xcite ) , @xmath14 kyr and @xmath15 g for the pulsar .
assuming that @xmath3 is similar to the pulsar s its true age , the transverse velocity implied by the pulsar s displacement from the apparent center of the g5.41.2 is greater than @xmath16 km s@xmath13 ( @xcite ; @xcite ) .
observations of psr b175724 taken six years apart , however , failed to detect the expected proper motion from the pulsar , yielding a distance - independent lower limit on the age of psr b175724/g5.41.2 of @xmath17 kyr ( @xcite ) .
this is more than a factor of two greater than the pulsar s mdr characteristic age .
the discrepancy between the proper motion age and the mdr characteristic age of psr b175724 suggests that the spin - down of the pulsar is not due purely to mdr but also has significant contributions from other sources of torque .
istomin ( 1994 ) considered a model for psr b175724 in which the pulsar was interacting with dense plasma in the shell of g5.41.2 , causing an increase in the torque at the light cylinder . here
we consider another possible source of extra torque on the neutron star from a disk of material accreted from ejecta produced in the supernova explosion .
these _ fallback disks _ may be roughly divided into two categories : `` prompt '' and `` delayed '' .
prompt disks may be formed from @xmath18 ( @xcite ; @xcite ) of ejecta material soon after the initial core collapse in a type ii supernova explosion ( @xcite ) .
formation of such prompt disks is probably limited to @xmath19 days after the core collapse because of heating of the ejecta by @xmath20ni decays ( @xcite ) .
delayed disks may form years after the explosion from ejecta decelerated by radiative cooling ( @xcite ) or by a strong reverse shock ( @xcite ) caused by the primary supernova blast wave impinging on dense circumstellar material from the pre - supernova stellar wind ( @xcite ) . whether or not a neutron star accretion disk can form shortly after a supernova explosion depends on the opposing forces of the pulsar mdr wind and the pressure of the hot and turbulent environment shortly after the explosion .
since the latter is highly uncertain ( @xcite ) , for the purposes of this _ letter _ we assume that a disk can form around a neutron star under these conditions and explore the implications for psr b175724 .
an accretion disk around a magnetized neutron star can exert a spin - down torque on the star if the mass inflow rate is low and the magnetic field is strong .
quantitatively , this condition is met when the keplerian co - rotation radius @xmath21 cm is less than the magnetospheric radius @xmath22 cm , where @xmath23 g s@xmath13 is the mass inflow rate onto the magnetosphere and @xmath24 g is the strength of the neutron star dipole magnetic field ( see e.g @xcite ) . here and elsewhere
we assume a neutron star mass and radius of @xmath25 and @xmath26 km , respectively . when @xmath27 , infalling material is stopped at the magnetosphere by a centrifugal barrier which prevents accretion onto the neutron star surface .
instead , the infalling material may be accelerated away in a wind which carries away angular momentum from the magnetosphere and hence the neutron star itself .
this `` propeller effect '' ( @xcite ) spin - down mechanism has been invoked to explain the behavior of some galactic accretion powered x
ray pulsars ( @xcite ; @xcite ) , the spin - evolution of anomalous x ray pulsars ( axps : @xcite ; @xcite ) and soft gamma
ray repeaters ( @xcite ; @xcite ) , and the optical and infrared spectra of some radio pulsars ( @xcite ) .
the pulsar b175724 is in the propeller regime ( @xmath28 ) for mass inflow rates of @xmath29 g s@xmath13 and a canonical ( @xcite ; @xcite ) neutron star magnetic field of @xmath30 g. since the characteristic age for an old ( @xmath31 kyr ) pulsar depends only weakly on the initial spin period ( @xcite ) , we neglect the different formation times of the delayed and prompt fallback disks , and assume that the spin - down evolution of the pulsar at all times is determined by the combined torque from both the mdr and the accretion disk
. the spin - down rate due to radiation from a rotating magnetic dipole is ( @xcite ) @xmath32 where @xmath33 is the neutron star moment of inertia .
the spin - down rate due to the propeller torque vanishes gradually ( @xcite ; @xcite ) as the star approaches spin - equilibrium ( @xmath34 ) at a spin period @xmath35 s. the spin - down rate due to the propeller torque is simply ( @xcite ) @xmath36 where @xmath37 is a positive constant of order unity ( @xcite ) and @xmath38 . assuming a constant mass inflow rate and dipole magnetic field , the timing age for a final spin period @xmath39 is given by @xmath40 where @xmath41 is the initial angular frequency . a more realistic expression for
the propeller torque would incorporate a time dependent mass inflow rate ( @xcite ) in equation ( [ wind ] ) , as @xmath42 should decrease in time as the disk dissipates .
for the model to be correct , however , a disk must still be present around the pulsar , because otherwise the mdr timing age would be greater than the true age .
therefore the @xmath42 used here may be thought of as a time - averaged value of the mass inflow rate .
in addition , the effect of the propeller flow on the mdr torque ( @xcite ) is not taken into account .
we plan on incorporating both of these effects in future work .
a contour plot of the pulsar b175724 timing age @xmath43 for various values of the magnetic field strength @xmath44 and mass inflow rate @xmath42 is show in figure 1 .
the characteristic ages were calculated using equations ( [ mdr][age ] ) and assuming @xmath11 s , @xmath45 ms , and @xmath46 . in this simple model ,
the allowed values of @xmath44 and @xmath42 for pulsar b175724 lie on the heavy solid line corresponding to the observed @xmath7 and @xmath8 .
in addition , the shaded regions of parameter space in figure 1 are excluded by the lower limit on the age ( right shaded region : @xcite ) , and the necessary condition @xmath27 ( shaded region in upper left corner ) . from figure 1 , we find that values of @xmath47 g , @xmath48 g s@xmath13 , and @xmath49 kyr are consistent with the lower limit on the true age of @xmath17 kyr ( @xcite ) and the present - day spin - down rate of the pulsar .
in the context of this model , the required mass inflow rate for psr b175724 overlaps the range of @xmath42 inferred from accretion powered neutron star systems ( @xcite ) .
radio pulsations are not seen from accretion powered x
ray pulsars in binary systems ( @xcite ) , which implies that the emission mechanism responsible for radio pulsations may be quenched by matter near the polar caps of the pulsar .
this is not a problem for the psr b175724 model , however , because in propeller sources most of the matter is ejected before it has a chance to reach the polar cap and quench the radio emission .
radio emission may be suppressed for propeller sources closer to spin equilibrium than psr b175724 ( e.g. the axps : @xcite ) , however , because as equilibrium is approached more matter will be allowed to fall onto the neutron star surface .
the total mass required to fuel the propeller spin - down over the lifetime ( so far ) of pulsar b175724 would be @xmath50 a tiny fraction of the total amount of ejecta in a typical type ii supernova explosion ( @xcite ) .
this hypothesis can be tested by multiwavelength observations . in this model ,
the total emission from the pulsar would be due to the propeller wind , mdr , and thermal emission from the accretion disk .
wang & robertson ( 1985 ) calculated the angle integrated thermal bremsstrahlung emissivity @xmath51 from the heated plasma in a propeller flow .
using their scalings , the x ray luminosity and temperature are given by @xmath52 erg s@xmath13 and @xmath53 kev , respectively , where @xmath54 is the width of the magnetospheric boundary layer where the plasma is heated by the magnetic field .
this assumes spherical symmetry , so the luminosity will probably be @xmath55 ergs s@xmath13 or less for the case of a disk geometry .
if psr 175724 is not a propeller source , we would expect the non - thermal x ray emission which is characteristic of young rotation
powered pulsars ( @xcite )
e.g. a power law emission spectrum with a photon index @xmath56 and an x ray luminosity given by the empirical relation @xmath57 erg s@xmath13 . at a distance of @xmath58 kpc ( @xcite ) ,
the flux from the non - thermal , non - propeller emission would be @xmath59 ergs @xmath60 s@xmath13 , which would be easily detectable by _ xmm _ or _
chandra_. the detection of dimmer , thermal x ray emission instead of the brighter non - thermal emission would be evidence in support of the propeller model
. cooler disk blackbody emission could also be visible at optical and infrared wavelengths from the accretion disk .
the spectrum of propeller disks depend on @xmath44 , @xmath42 , and the disk orientation with respect to the line of sight ( e.g. @xcite ) , but we can place an upper limit on the optical emission from a psr b175724 disk in the following manner . assuming that the majority of the optical disk emission originates at the innermost disk radii ( @xcite ) , the upper limit on the optical luminosity is @xmath61 @xmath62 .
this estimate ignores heating of the disk by irradiation from the pulsar , which dominates the heating only for pulsar luminosities @xmath63 ergs s@xmath13 ( @xcite ) .
assuming @xmath64 , @xmath65 , and an apparent visual magnitude of @xmath66 for the sun ( @xcite ) , the apparent magnitude of the disk at @xmath58 kpc ( uncorrected for extinction ) is @xmath67 .
the extinction @xmath68 can be estimated from the formula @xmath69 mag @xmath60 ( @xcite ) , where @xmath70 is the @xmath71 column density along the line of sight . at a distance of @xmath58 kpc , @xmath72 @xmath60 in the galactic plane ( as can be seen from the axp spectral data in @xcite ) , which yields an extinction of @xmath73 mag and a lower limit on the disk magnitude of @xmath74 .
this is comparable to the estimated magnitudes of propeller disks around axps as calculated by perna et al .
( 2000 ) .
if psr b175724 is indeed surrounded by an accretion disk , then fossil accretion disks and propeller spin - down may be present , or may have been present at one time , in a significant fraction of isolated pulsars .
this would affect the distributions of pulsar ages , magnetic field strengths , and space velocities inferred using the pulsar s @xmath7 and @xmath8 values and the mdr formulae discussed in @xmath75 .
magnetic field strengths estimated using equation ( [ eq0 ] ) , for example , would overestimate the true field strengths for both pulsars currently undergoing propeller spin - down _ and _ for pulsars which had experienced some propeller spin - down in the past ( and whose disks had subsequently dissipated ) .
the effect on the distribution of radio pulsar ages and hence the distribution of pulsar velocities inferred from their angular positions ( e.g. @xcite ) depends on whether each pulsar is currently undergoing propeller spin - down or not .
if a pulsar is currently undergoing propeller spin - down , then the mdr timing age @xmath3 is an _ underestimate _ of the true age ( as exemplified by psr b175724 ) .
if a pulsar had a propeller disk which then dissipated , its present day mdr age would be an _ overestimate _ of its true age , since the pulsar had undergone a period of increased spin - down rate ( over the mdr spin - down rate ) in the past .
the latter scenario is probably more prevalent in the observed population of radio pulsars , since radio pulsations may be scattered and quenched in neutron stars with strong propeller flows ( e.g. @xcite ; @xcite ) , and therefore one might expect a systematic overestimation of radio pulsar ages due to fossil disks and propeller spin - down .
curiously , such a systematic overestimation of pulsar ages using mdr has been inferred from some pulsar population studies ( @xcite ) .
we have shown that the addition of torques from an accretion disk can explain the discrepancy between the mdr timing age of psr b175724 and its true age .
this model can be tested through x ray and optical observations of this pulsar . the accretion disk model for pulsar b175724 leaves
open the question of whether or not pulsar b175724 is an unusual and rare object or if it instead reflects a generic feature in the evolution of neutron stars .
if the former is true , the accretion model removes the need ( @xcite ) to revise our current understanding of the physics and astrophysics of neutron stars because of this single pulsar . if the latter is true , however , the distributions of pulsar magnetic field strengths , ages , and velocities will have to be reconsidered to take into account the effects of increased spin - down due to accretion disk torques .
observations of young pulsars associated with supernova remnants may hold the key towards resolving this question , because the pulsar ages can be constrained independently of the pulsar temporal parameters .
we thank the referee for helpful comments .
this work was performed while one of the authors ( dm ) held a national research council - gsfc research associateship .
rer acknowledges support by nasa contract nas5 - 30720 , and rel support from the astrophysical theory program . | radio pulsars are thought to spin - down primarily due to torque from magnetic dipole radiation ( mdr ) emitted by the time - varying stellar magnetic field as the star rotates .
this assumption yields a ` characteristic age ' for a pulsar which has generally been assumed to be comparable to the actual age .
recent observational limits on the proper motion of pulsar b175724 , however , revealed that the actual age ( @xmath0 kyr ) of this pulsar is much greater than its mdr characteristic age ( @xmath1 kyr ) calling into question the assumption of pure mdr spin - down for this and other pulsars . to explore the possible cause of this discrepancy
, we consider a scenario in which the pulsar acquired an accretion disk from supernova ejecta , and the subsequent spin - down occurred under the combined action of mdr and accretion torques .
a simplified model of the accretion torque involving a constant mass inflow rate at the pulsar magnetosphere can explain the age and period derivative of the pulsar for reasonable values of the pulsar magnetic field and inflow rate .
we discuss testable predictions of this model . |
spin properties attract the great attention in recent years due to attempts to realize an electronic device based on the spin of carriers .
conduction electrons are obvious candidates for such devices , particularly in nanostructures where electron energy spectrum and shape of the envelope functions can be effectively engineered by the growth design , application of electric or magnetic fields as well as by illumination with light .
various semiconductor materials are being involved in the spintronics activities .
sige quantum well structures are among them .
although bulk si and ge have a center of symmetry , quantum - well structures grown from these materials can lack such a center and allow the spin splitting of the electronic subbands .
the quantum engineering of spintronics devices is usually focused on the rashba spin - dependent contribution to the electron effective hamiltonian in heterostructures .
this contribution appears due asymmetry of the heterostructure ( the so - called structure inversion asymmetry , or sia ) and has no relation to the properties of a bulk semiconductor .
it is commonly believed that the dresselhaus contribution caused by bulk inversion asymmetry ( bia ) is absent in structures grown from centrosymmetric materials . in the present work
we show for the first time that the dresselhaus - like spin splitting is possible in heterostructures made of si and ge , as an alternative to the rashba effect , and it occurs due to the anisotropy of chemical bonds at interfaces .
we consider quantum well ( qw ) structures grown along the axis @xmath1 $ ] , and introduce two orthogonal directions in the interface plane , @xmath2 $ ] and @xmath3 $ ] . depending on the properties of an interface between si@xmath4ge@xmath5 and si , its symmetry on average can be c@xmath6 or @xmath7 .
@xcite the former point group describes the symmetry of an ideal heterointerface with the interfacial chemical bonds lying in the same plane .
a nonideal interface containing monoatomic fluctuations has two kinds of flat areas with interfacial planes shifted with respect to each other by a quarter of the lattice constant .
the local symmetry of each area is c@xmath8 as well .
however if the both kinds are equally distributed , the interface overall symmetry increases up to @xmath7 .
it follows then that the symmetry of a si@xmath4ge@xmath5/si quantum well structure containing two interfaces is described by one of five point groups : @xmath9 or @xmath10 in case of two ideal interfaces with odd or even number of monolayers between them ; @xmath11 for a pair of ideal and rough interfaces ; @xmath7 or @xmath12 for two non - ideal interfaces of the overall symmetry @xmath7
each | see fig . 1 . = = 0.3 the lowest conduction band in ( 001)-sige / si qws is located near the @xmath13 point of the brillouin zone . at this point ,
the bulk electron states form projective representations of the point group @xmath12 .
all five above - mentioned groups are subgroups of @xmath12 .
two of them , @xmath12 and @xmath10 , contains the space inversion operation and forbid the spin splitting of electronic states .
three remaining groups allow the spin - dependent linear - in-@xmath0 term @xmath14 . here
@xmath15 are the pauli matrices and @xmath16 is the in - plane electron wave vector . note that in the following we consider the electronic states attached to the @xmath17 valley which lies lower than @xmath18 and @xmath19 valleys .
@xcite in the group @xmath9 , the dresselhaus - like term @xmath20 is an only invariant which can be constructed from the products @xmath21 . in particular ,
the rashba term @xmath22 is forbidden . on the contrary , in the group @xmath7
, the only invariant combination is @xmath22 .
the analysis shows that both combinations @xmath23 and @xmath24 are invariants of the group @xmath11 .
hence the latter allows both dresselhaus- and rashba - like spin - dependent terms .
microscopically , a linear - in-@xmath25 correction to the conduction - band effective hamiltonian is given by the second order of the perturbation theory @xmath26 here @xmath27 and @xmath28 are the @xmath29 and spin - orbit interaction hamiltonians , respectively .
we take into consideration only the coupling of the conduction states @xmath30 with the valence states @xmath31 , where the indices @xmath32 enumerate the degenerate states .
the symmetry properties of the bloch functions at the @xmath13 point of a diamond - lattice semiconductor crystal coincide with those of the following functions @xcite @xmath33 @xmath34 where @xmath35 is the lattice constant . for the bulk states @xmath36 in the bases ( [ x1 ] ) , ( [ x4 ] ) ,
the interband matrix elements can be presented as @xmath37 here @xmath38 , @xmath39 , @xmath40 is the momentum operator , and @xmath41 is the @xmath42-component of the pseudovector @xmath43 which enters into the spin - orbit hamiltonian @xmath44 .
the substitution of ( [ kpso ] ) into ( [ second_order ] ) gives a spin - dependent matrix @xmath45 proportional to @xmath46 which however _ does not _ lead to removal of the degeneracy of the @xmath47-states in the bulk centrosymmetric material as expected .
the splitting can be achieved if one takes into account the anisotropy of chemical bonds at the interfaces which results in @xmath48-functional contributions to @xmath27 and @xmath49 of the form @xmath50 where @xmath51 is the interface coordinate , and the matrices @xmath52 have few linearly independent components . in the case of the lowest symmetry under study , @xmath11 , each matrix
is determined by four linearly - independent parameters @xmath53 in the bases ( [ x1 ] ) , ( [ x4 ] ) , the parameters @xmath54 and @xmath55 @xmath56 are purely imaginary . according to eq .
( [ second_order ] ) , the linear in @xmath57 correction to the conduction - band hamiltonian for a single interface of the @xmath11-symmetry , is given by @xmath58 where @xmath59 is a linear combination of the pauli matrices .
treating the perturbation ( [ deltakpso ] ) in the first - order approximation , one can obtain @xmath60 \nonumber\\ & - & { { \cal u } \over e_0 } \biggl [ ( \sigma_x k_x - \sigma_y k_y ) ( p_1 + p_3 ) + ( \sigma_x k_y - \sigma_y k_x ) ( p_2 + p_4 ) \biggr ] , \nonumber\end{aligned}\ ] ] where @xmath61 is the energy separation between @xmath47 and @xmath62 states .
the electron effective hamiltonian in an ideal qw contains a sum of two contributions ( [ c2vif ] ) related to the left- and right - hand - side interfaces .
the corresponding parameters @xmath63 or @xmath64 are interconnected by @xmath65_r = [ u_3 , u_4 , u_1 , u_2]_l,\ ] ] @xmath66_r = [ - p_3 , - p_4 , - p_1 , -p_2]_l,\ ] ] if the qw contains an even number of monoatomic layers ( @xmath67 symmetry ) , and @xmath68_r = [ u_1 , - u_2 , u_3 , - u_4]_l,\ ] ] @xmath66_r = [ p_1 , -p_2 , p_3 , -p_4]_l,\ ] ] if the qw contains an odd number of monoatomic layers ( @xmath69 symmetry ) .
it readily follows from here that , in the former case , the electronic states in si / ge qws are spin - degenerate ( @xmath70 ) and , in the latter , the spin degeneracy is removed by a dresselhaus - like linear - in-@xmath0 terms .
the similar analysis can be carried out for non - ideal si / ge qw structures of the @xmath11 and @xmath7 symmetry .
in this work , we have analyzed spin splitting of electron states in si / ge heterostructures of all possible symmetries .
this opens a possibility to discuss the recent spin relaxation times measurements in these heterostructures .
@xcite the work on estimation of the interface parameters @xmath55 and @xmath54 in the microscopic tight - binding model is in progress .
this work is financially supported by the rfbr , dfg , intas , by the programmes of russian ministry of science and presidium of ras , and by the dynasty foundation and icfpm .
ganichev , u. rssler , w. prettl , e.l .
ivchenko , v.v .
belkov , r. neumann , k. brunner , and g. abstreiter , phys .
b * 66 * , 75328 ( 2002 ) .
z. wilamowski , w. jantsch , h. malissa , and u. rssler , phys .
b * 66 * , 195315 ( 2002 ) .
xudong xiao , chun zhang , a.b .
fedorov , zhenghao chen , and m.m .
loy , j. vac .
b * 15 * , 1112 ( 1997 ) .
v.ya . aleshkin and n.a .
bekin , j. phys . : condens .
matter * 9 * , 4841 ( 1997 ) .
peter y. yu and manuel cardona , _ fundamentals of semiconductors .
physics and material properties _ , springer - verlag , berlin , heidelberg , 1996 . | spin splitting of conduction electron states has been analyzed for all possible point symmetries of sige quantum well structures .
a particular attention is paid to removal of spin degeneracy caused by the rotoinversion asymmetry of a ( 001 ) heterointerface between two diamond - lattice materials .
consequences of the spin splitting on the electron spin relaxation time is discussed . |
the coulomb gauge is one of the most popular gauges in quantum chromodynamics ( qcd ) .
in addition to the convenient choice for calculations , the coulomb - gauge qcd is also interesting from the theoretical point of view , especially in the context of the canonical quantization @xcite .
in addition , color confinement has been investigated in the coulomb gauge in terms of the color - coulomb interaction @xcite . as a new picture of hadrons
, the gluon - chain picture also stems from the coulomb - gauge qcd @xcite . in the coulomb - gauge qcd
, the faddeev - popov ( fp ) operator is the important key quantity . actually , the color - coulomb energy is enhanced by the near - zero fp eigenmodes , which would lead to the confining force between color charges .
this confinement scenario is known as the gribov - zwanziger scenario @xcite . from lattice - qcd numerical calculations ,
the color - coulomb energy certainly gives a linear interquark potential @xcite .
the confinement is also characterized by the infinite self - energy of an isolated color - charge . in the coulomb gauge
, the color - coulomb self - energy is expressed in terms of the fp eigenmodes , and its divergence is originated from near - zero fp eigenmodes , as was shown in lattice qcd calculations @xcite .
then , it is meaningful to investigate the fp eigenmodes for the understanding of non - perturbative qcd @xcite . in this paper
, we investigate the fp eigenmodes in terms of the gluon - momentum components .
since the qcd interactions are mediated by gluons , the fp eigenmodes should reflect the properties of the gluon . to analyze the correspondence between the gluon field and the fp eigenmodes ,
we decompose the link - variables in terms of momentum components , and remove infrared or ultraviolet momentum modes in lattice qcd @xcite .
the organization of this paper is as follows . in sec.ii , we review the color - coulomb energy and properties of the fp eigenmodes in the coulomb gauge .
we also introduce the gluon - momentum cut method in lattice qcd . in sec.iii
, we show the lattice qcd results for the color - coulomb energy and the fp eigenmodes with the ir / uv - momentum gluon cut .
section iv will be devoted to summary and discussions .
in this section , we review the color - coulomb energy and the fp operator properties in the coulomb gauge @xcite
. we also briefly introduce the gluon - momentum cut method in lattice qcd .
since we discuss the coulomb - gauge qcd at a fixed timeslice @xmath0 , we will omit the time coordinate @xmath0 in this paper .
the coulomb gauge is one of the most popular gauges in both analytic framework and lattice qcd calculation .
the definition of the coulomb gauge is given by @xmath1 where the gluon fields @xmath2 with generator @xmath3 ( @xmath4 ) .
the coulomb gauge is also defined by the minimization of the global quantity @xmath5 by the gauge transformation .
the minimization of the quantity @xmath6 means that the spatial gauge - field fluctuations are maximally suppressed in the coulomb gauge .
these gauge - fixing conditions resemble the landau - gauge condition , although the lorentz covariance is partially broken . in the coulomb gauge
, gluon fields are decomposed into the canonical variable @xmath7 and the potential @xmath8 @xcite .
one of the advantages in the coulomb gauge is the compatibility with the canonical quantization @xcite . in the coulomb gauge ,
the qcd hamiltonian is expressed with color electric and magnetic fields , @xmath9 and @xmath10 , as @xmath11 where @xmath12 is the color - charge density , and @xmath13 the instantaneous coulomb propagator .
@xmath13 is defined as @xmath14_{\vec{x}\vec{y}}^{ab},\ ] ] using the fp operator @xmath15 with @xmath16 .
the second term of the hamiltonian ( [ eq : qcdhamiltonian ] ) corresponds to the color - coulomb energy , which is expressed as @xmath17_{\vec{x}\vec{y}}^{aa } \rangle \label{eqvcoul}\ ] ] with @xmath18 , the quadratic casimir @xmath19 of the fundamental representation , and the coupling constant @xmath20 @xcite . in the abelian gauge theory ,
the fp operator becomes the laplacian , and the instantaneous coulomb propagator is inverse of the laplacian .
thus , the coulomb energy becomes familiar coulomb potential form @xmath21 . in the non - abelian gauge theory ,
the fp operator has non - trivial zero - modes , which form the gribov horizon @xcite . in the neighborhood of the gribov horizon ,
the color - coulomb energy is largely enhanced , which is conjectured to contribute to the confining force .
this confinement picture is known as the `` gribov - zwanziger scenario '' @xcite . instead of the direct definition of the color - coulomb energy @xmath22 in eq.([eqvcoul ] ) , we consider the correlator of the time - like wilson lines as @xmath23 \rangle,\ ] ] with @xmath24 and the time - like wilson line @xmath25 as for the relation to the color - coulomb energy , one finds @xcite @xmath26 in lattice qcd , the qcd action is constructed from the link - variable @xmath27 , which is defined as @xmath28 with the lattice spacing @xmath29 and the gauge coupling constant @xmath20 @xcite .
the coulomb - gauge fixing condition is expressed in terms of the link - variable , and is given by the maximization of @xmath30 \equiv \sum_{\vec{x } } \sum_{i=1}^3 \mathrm{re } \
\mathrm{tr } \
u_i(\vec{x } ) , \label{eq : r_coul}\ ] ] by the gauge transformation @xmath31 with @xmath32 . in the continuum limit @xmath33
, this condition results in the minimization of eq.([eqcoulombglobal ] ) . using the time - like wilson line correlator @xmath34 on lattice
, we define @xmath35 as @xmath36.\ ] ] especially , at @xmath37 , we call @xmath38 as `` instantaneous potential '' @xcite , which is considered to be closely related to the color - coulomb energy in eq.([eqvcoul ] ) . actually , in the continuum limit , the instantaneous potential would coincide with the color - coulomb energy as in eq.([eqvcouldef2 ] ) @xcite .
the lattice qcd calculations show that the instantaneous potential gives a linear rising potential @xcite , and satisfies the casimir scaling @xcite similar to the physical interquark potential @xcite .
however , the slope of the potential is @xmath39 times larger than physical string tension @xmath40gev / fm @xcite .
the color - coulomb energy is directly calculated based on eq.([eqvcoul ] ) in both su(2 ) @xcite and su(3 ) @xcite lattice qcd .
these lattice qcd calculations also indicate the overconfining potential with a larger string tension .
actually , the color - coulomb energy gives an upper bound on the static interquark potential @xmath41 , @xmath42 which was shown by zwanziger @xcite .
this large color - coulomb energy is considered as the overconfining state , and the gluon - chain picture is proposed for the true ground state of the quark - antiquark system in the coulomb gauge @xcite . in spite of overconfining
, the color - coulomb energy is expected to relate to the confinement at least at zero temperature .
therefore , we concentrate on the color - coulomb energy properties in this paper . in the coulomb - gauge lattice qcd , the fp operator is given by @xmath43 \end{aligned}\ ] ] using the generators @xmath3 and link - variable @xmath27 .
on @xmath44 lattice , the total number of fp eigenmodes is @xmath45 with the spatial volume @xmath46 .
the fp operator @xmath47 has trivial @xmath48 zero - modes @xmath49 with @xmath50 . in free - field and qed cases ,
the fp operator becomes the laplacian , and the eigenvalues are expressed by the three - dimensional momentum @xmath51 . in the lattice theory ,
the eigenvalue is given by @xmath52 with @xmath53 $ ] .
in the continuum limit , it becomes @xmath54 in terms of the fp eigenfunction , the relation between the color - coulomb energy and the gribov horizon becomes clear .
considering the fp eigenstate @xmath55 which satisfies @xmath56 with eigenvalue @xmath57 , and the fp eigenfunction is given by @xmath58 where @xmath29 is the color index .
using the fp eigenmodes , the color - coulomb energy in eq.([eqvcoul ] ) is expressed as @xmath59 with @xmath18 @xcite .
equation ( [ eq : vcoulfpmode ] ) indicates that low - lying fp eigenmodes would give dominant contribution to the color - coulomb energy . from lattice
qcd calculations , the color - coulomb energy is brought by only small number of low - lying fp eigenmodes @xcite .
the confinement is also investigated in terms of the color - coulomb self - energy , since an isolated color - charge has infinite energy in the infrared in qcd . using the fp eigenmodes ,
the color - coulomb self - energy @xcite is expressed as @xmath60 @xmath61 is the diagonal - matrix element of the laplacian operator : @xmath62 \psi_n^a(\vec{y } ) .
\end{aligned}\ ] ] here , we define the eigenmode density @xmath63 as @xmath64 where @xmath65 is the number of eigenvalues in @xmath66 $ ] . in the infinite - volume limit
, color - coulomb self - energy @xmath67 @xcite is expressed as @xmath68 with the fp - eigenmode density @xmath63 , and the uv cutoff @xmath69 .
therefore , if the criterion @xmath70 is satisfied , color - coulomb self - energy @xmath67 diverges in the infrared limit @xcite .
this criterion indicates the importance of the near - zero fp modes and matrix elements @xmath71 for confinement , which is similar to the dirac zero - mode and chiral symmetry breaking in the banks - casher relation @xcite . in free - field and qed cases ,
the fp eigenvalue is given by eq.([eq : fpeigenfreecont ] ) , and then @xmath63 and @xmath71 behave as @xmath72 therefore , the confinement criterion is not satisfied as @xmath73 and the coulomb self - energy is infrared finite .
since the color - charge is confined in qcd , color - coulomb self - energy is expected to diverge at infrared limit . from su(2 )
lattice - qcd analysis , the fp spectrum @xmath63 and the matrix element @xmath71 approximately behave as @xmath74 near the horizon @xmath75 @xcite .
then the divergence criterion is satisfied as @xmath76 therefore , the color - coulomb self - energy actually diverges in the infrared limit and the color - charge is confined in the non - abelian case . in this subsection
, we introduce the formalism to remove gluon - momentum modes in lattice qcd , which proposed in ref.@xcite . here
, we use the coulomb gauge , and carry out the three - dimensional fourier transformation with respect to the spatial coordinate @xmath77 at fixed timeslice @xmath0 .
the procedure is as follows .
we generate the link - variable @xmath78 on @xmath44 lattice with lattice spacing @xmath29 , and fix the coulomb gauge .
we carry out the discrete fourier transformation . the momentum - space link - variable is given by @xmath79 with the spatial volume @xmath46 , and momentum @xmath53 $ ] . in the momentum space
, the lattice spacing @xmath80 is given by @xmath81 \3 .
we introduce infrared and ultraviolet `` cut '' @xmath82 $ ] , and replace the link - variable by the free - field link - variable as @xmath83 the fourier transformation of the free - field link - variable @xmath84 is @xmath85 \4 . in order to return to the coordinate space variable
, we carry out the inverse fourier transformation of @xmath86 as @xmath87 since this link - variable @xmath88 is not su(@xmath89 ) matrix , we project it onto su(@xmath89 ) matrix @xmath90 by maximizing @xmath91,\ ] ] which is often used in lattice qcd algorithm .
finally , we obtain the momentum projected link - variable @xmath92 .
in this section , we analyze the instantaneous potential , the fp eigenmodes , and the color - coulomb energy , in terms of the gluonic momentum using su(3 ) lattice monte carlo calculations .
we mainly use @xmath93 lattice at @xmath94 , which corresponds to the lattice spacing @xmath95fm @xcite , and @xmath96gev in the momentum space .
we use the jack - knife method for the estimating of the statistical error .
first , we analyze the instantaneous potential with the ir / uv - momentum gluon cut . using @xmath97 ,
we define the instantaneous potential with gluon - momentum cut as @xmath98 with @xmath24 .
figure [ figcolorcoulomb](a ) and ( b ) are @xmath99 with the ir / uv - momentum cut , respectively .
we also show original ( no momentum cut ) instantaneous potential , and fit result using coulomb plus linear form .
the best fit value of the slope is @xmath100 . as shown in fig.[figcolorcoulomb](a ) , the instantaneous potential becomes non - confining with ir - momentum cut .
the instantaneous potential changes drastically even for the smallest ir - cut .
in contrast to the ir - cut , the instantaneous potential is almost unchanged by the uv - cut as shown in fig.[figcolorcoulomb](b ) . the instantaneous potential @xmath99 with the ir / uv - momentum gluon cut for @xmath93 lattice with @xmath95fm , i.e. , @xmath96gev .
the original ( no momentum cut ) instantaneous potential is added with the fitting curve of coulomb plus linear form .
the statistical error is small and the error bars are hidden in the symbols .
( a ) ir - momentum cut with @xmath101 and @xmath102 . ( b ) uv - momentum cut with @xmath103 and @xmath104 .
an irrelevant constant is shifted.,title="fig:",scaledwidth=50.0% ] the instantaneous potential @xmath99 with the ir / uv - momentum gluon cut for @xmath93 lattice with @xmath95fm , i.e. , @xmath96gev .
the original ( no momentum cut ) instantaneous potential is added with the fitting curve of coulomb plus linear form .
the statistical error is small and the error bars are hidden in the symbols .
( a ) ir - momentum cut with @xmath101 and @xmath102 .
( b ) uv - momentum cut with @xmath103 and @xmath104 . an irrelevant constant is shifted.,title="fig:",scaledwidth=50.0% ] next , we analyze the fp eigenmodes with the ir / uv - momentum gluon cut .
the instantaneous potential becomes non - confining with the ir - cut .
similar to the color - coulomb energy , the instantaneous potential would be closely related to the fp eigenmodes , we expect that the fp eigenmodes are largely changed with the ir - cut .
first , we evaluate the low - lying 250 fp eigenmodes using arpack @xcite . in this case , the total number of fp eigenmodes is @xmath105 .
figure [ figfpspectrumlow](a ) and ( b ) are the low - lying fp spectrum @xmath63 with the ir / uv - momentum gluon cut , respectively . in both figures
, we have added the original fp spectrum for comparison , and we have omitted the trivial 8 zero - modes in eq.([eq : trivialzero ] ) , which always remain even after momentum cut procedure . in fig.[figfpspectrumlow](a ) , the vertical bars denote the free - field eigenvalues @xmath106 and @xmath107 .
low - lying 250 fp spectrum @xmath63 with the ir / uv momentum cut and original spectrum for @xmath93 lattice with @xmath95fm , i.e. , @xmath96gev .
the binwidth is taken as @xmath108 .
( a ) ir - momentum cut with @xmath109 , and @xmath110 .
the vertical bars denote the non - zero free - field spectrum .
( b ) uv - momentum cut with @xmath103 and @xmath104 .
, title="fig:",scaledwidth=50.0% ] low - lying 250 fp spectrum @xmath63 with the ir / uv momentum cut and original spectrum for @xmath93 lattice with @xmath95fm , i.e. , @xmath96gev .
the binwidth is taken as @xmath108 .
( a ) ir - momentum cut with @xmath109 , and @xmath110 .
the vertical bars denote the non - zero free - field spectrum .
( b ) uv - momentum cut with @xmath103 and @xmath104 .
, title="fig:",scaledwidth=50.0% ] as shown in fig.[figfpspectrumlow](a ) , the ir - cut fp spectrum is drastically changed .
the near - zero fp modes vanish , and the spectrum changes into multi - peak structure from original smooth one . by increasing the ir - cut @xmath111 ,
the peaks of the spectrum become sharper , and the fp spectrum tends to converge into @xmath112-functional peaks in the free - field limit as eq.([eq : fpeigenfree ] ) . on the other hand ,
the fp spectrum is almost unchanged for the uv - cut .
this uv - cut insensitivity is the same as the instantaneous potential as shown in fig.[figcolorcoulomb](b ) .
next , we analyze the matrix element of the laplacian operator , i.e. , @xmath113 , which is important for both color - coulomb energy in eq.([eq : vcoulfpmode ] ) and self - energy divergence condition ( [ eqselfenergydivergence ] ) . here , we mainly discuss the diagonal component @xmath114 , since the off - diagonal elements @xmath113 are found to be almost the zero from lattice qcd calculations .
figure [ figflambda ] is the scatter plot of the diagonal element @xmath71 with the ir / uv - momentum cut , for low - lying 250 eingenmodes .
we also show the original @xmath71 without momentum cut , and omit the trivial 8 zero - modes in this figure . in fig.[figflambda](a )
the solid - box symbols denote the free - field values @xmath106 and @xmath107 in this @xmath93 lattice , and the solid line denotes free - field value @xmath115 in the continuum theory . by the ir - momentum cut , near - zero modes of @xmath71 vanish as in the ir - cut fp spectrum .
@xmath71 changes into band - like structure from original smooth distribution , and tends to converge into free - field value on lattice . in fig.[figflambda ] , one finds a `` flow '' of eigenmodes into free - field limit by the ir - momentum cut . since both matrix element @xmath71 and fp spectrum @xmath63 converge to free - field limit , the self - energy divergence condition ( [ eqselfenergydivergence ] ) would not be satisfied by the ir - cut . actually , the instantaneous potential becomes non - confining as shown in fig.[figcolorcoulomb](a ) . on the other hand
, @xmath71 is almost unchanged by the uv - momentum cut like the fp spectrum .
the scatter plot of the diagonal matrix element @xmath114 with the ir / uv momentum cut , and original matrix element , for @xmath93 lattice with @xmath95fm , i.e. , @xmath96gev .
( a ) ir - momentum cut with @xmath109 , and @xmath110 .
the solid - box symbols denote for non - zero free - field values , and the solid line for free - field value in continuum limit , i.e. , @xmath115 .
( b ) uv - momentum cut with @xmath103 and @xmath104 .
, title="fig:",scaledwidth=50.0% ] the scatter plot of the diagonal matrix element @xmath114 with the ir / uv momentum cut , and original matrix element , for @xmath93 lattice with @xmath95fm , i.e. , @xmath96gev .
( a ) ir - momentum cut with @xmath109 , and @xmath110 .
the solid - box symbols denote for non - zero free - field values , and the solid line for free - field value in continuum limit , i.e. , @xmath115 .
( b ) uv - momentum cut with @xmath103 and @xmath104 .
, title="fig:",scaledwidth=50.0% ] as shown above , the low - lying fp eigenmodes are largely changed by the ir - momentum cut .
therefore , one may expect naively that low / high fp eigenmodes are closely related to the ir / uv - momentum gluon , respectively . to investigate correspondence between gluonic momentum and the fp spectrum , it is meaningful to calculate full fp eigenmodes .
however , it requires huge computational costs to perform full diagonalization of the fp operator for large volume lattices . here
, we adopt @xmath116 lattice with @xmath117 , which corresponds to the lattice spacing @xmath118fm and @xmath119gev @xcite .
the total number of the fp eigenmodes is @xmath120 .
we evaluate the full fp eigenmodes using lapack @xcite .
we have confirmed that the instantaneous potential and momentum cut dependence in this @xmath116 lattice , and find that qualitatively the same results are obtained in @xmath93 lattice with @xmath121 .
therefore , full - modes analysis is workable in this @xmath116 lattice .
figure [ figfpspectrumir ] and [ figfpspectrumuv ] are the full fp spectrum with the ir / uv momentum cut , respectively .
the solid curve denotes original fp spectrum in figs.[figfpspectrumir ] and [ figfpspectrumuv ] , and fig.[figfpspectrumir](d ) is free - field spectrum in this lattice size . for the ir - cut ,
the fp spectrum is drastically changed from smooth one to multi - peak structure in the whole eigenvalue region .
it is notable that both low and high fp eigenmodes are affected by the low momentum gluon .
the ir - cut spectrum clearly approaches to free - field spectrum , which is given by eq.([eq : fpeigenfree ] ) .
in contrast , the fp spectrum is almost unchanged for uv - cut as shown in fig.[figfpspectrumuv ] .
therefore , there are no direct correspondence between the ir / uv gluonic momentum and low / high fp eigenmodes , respectively . figures [ figfpspectrumir ] and [ figfpspectrumuv ] indicate that the smooth structure of fp eigenmodes mainly originate from the low - momentum gluon . as a caution , the uv - cut spectrum in fig.[figfpspectrumuv ] seems to resemble the ir - cut one at @xmath101 as shown in fig.[figfpspectrumir](a ) . however , as we discussed the low - lying fp eigenmodes , the near - zero modes vanish in the ir - cut , which differs from the uv - cut spectrum .
the near - zero eigenmodes are relevant for color - coulomb energy , such a small difference of low - lying spectrum can affect on the confining property of the color - coulomb energy . the full fp spectrum @xmath63 with the uv - momentum cut at @xmath122 for @xmath116 lattice at @xmath117 , i.e. , @xmath118fm and @xmath123gev . the binwidth is taken as @xmath124 .
the solid curve denotes original spectrum .
, scaledwidth=49.0% ] in sec.iii-a , we investigate the instantaneous potential instead of the color - coulomb energy . in this subsection
, we directly analyze the color - coulomb energy in eq.([eq : vcoulfpmode ] ) using the fp eigenmodes @xcite , which are evaluated in sec.iii-b .
figure [ figfpspectrumsum ] shows the color - coulomb energy with ir / uv - momentum cut using low - lying 50 fp eigenmodes apart from trivial 8 zero - modes .
for comparison , we add the original ( no momentum cut ) color - coulomb energy , and the instantaneous potential with the fitting curve of coulomb plus linear form . here , the color - coulomb energy is calculated directly with eq.([eq : vcoulfpmode ] ) . as shown in figs.[figcolorcoulomb ] and [ figfpspectrumsum ] , the instantaneous potential and the color - coulomb energy exhibit the similar momentum - cut dependence .
the uv - cut color - coulomb energy is almost the same as the original one , and the ir - cut energy becomes non - confining .
the color - coulomb energy @xmath22 obtained with eq.([eq : vcoulfpmode ] ) using low - lying 50 fp eigenmodes apart from trivial zero - modes .
for @xmath93 lattice at @xmath121 , i.e. , @xmath125fm and @xmath96gev , we show the ir - momentum cut with @xmath101 , the uv - momentum cut with @xmath103 , and original color - coulomb energy . for comparison
, the instantaneous potential is also added with the fitting curve of coulomb plus linear form .
an irrelevant constant is shifted .
, scaledwidth=50.0% ] in this paper , we have investigated the instantaneous potential and fp eigenmodes in terms of gluon - momentum components . in the context of the confinement mechanism
, there is an interesting similarity between our results on the ir - momentum gluon cut and those on the center - vortex removal @xcite , and then we compare these two different operations including their results in this subsection . the center vortex is an interesting object appearing in the maximal center ( mc ) gauge @xcite , and closely relates to the confinement @xcite and the infrared properties of qcd @xcite .
actually , when the center vortex is removed from the qcd vacuum , the string tension obtained from the wilson loop vanishes and the system becomes non - confining @xcite , which is observed in the ir - momentum cut of gluons @xcite .
also , as greensite and olejnk pointed out , by removing the center vortex in the mc gauge , the instantaneous potential in the coulomb gauge becomes non - confining @xcite , which resembles the ir - cut of gluons , as shown in fig.[figcolorcoulomb](a ) .
since the instantaneous potential is closely related to the color - coulomb energy , the non - confining instantaneous potential reflects the change of the near - zero fp eigenmodes . in the center - vortex removal ,
the fp spectrum becomes multi - peaks , and matrix element @xmath126 becomes band - like structure @xcite .
these changes are quite similar to those of the ir - momentum cut of gluons , as shown in figs.[figfpspectrumlow](a ) , [ figflambda](a ) , and [ figfpspectrumir](a)-(c ) .
in fact , both operations lead to the similar drastic change of fp eigenmodes . in the center - vortex removal , however , the fp spectrum does not coincide with the @xmath112-functional form of the free - field case , but becomes multi - peaks with a finite width .
then , there still remains the significant difference between `` no - vortex '' link - variables and free - field variables . in the gluon - momentum cut method , one can continuously change the link - variable to free - field one , and accordingly , the fp spectrum continuously goes from a smooth function to the @xmath112-functional form .
in fact , unlike the center - vortex removal , our operation is continuous . in the momentum - cut method , according to the ir cut , we observe a `` flow '' of the matrix element @xmath71 , which continuously goes to the free - field limit , as shown in fig.[figflambda](a ) . from the similarity between these two different operations ,
it is also interesting to investigate the correspondence between the center vortex in the mc gauge and the gluon - momentum components in the landau / coulomb gauge .
in this paper , we have investigated the relation between the fp eigenmodes and gluon - momentum components in the coulomb gauge using su(3 ) lattice qcd calculations at the quenched level .
the fp eigenmodes are considered to be important for confinement scenario in the coulomb gauge . especially , the low - lying eigenmodes lead the large color - coulomb energy , and the color - coulomb self - energy diverges from the enhancement of the near - zero eigenmode density .
we have analyzed low - lying fp eigenmodes with the ir / uv - momentum gluon cut , and also performed the full fp eigenmodes analysis . in the uv - momentum gluon cut , both color - coulomb energy and fp eigenmodes are almost unchanged , which indicates that high - momentum gluons are irrelevant for confining scenario in the coulomb gauge .
in contrast , the color - coulomb energy becomes non - confining , and the fp eigenmodes are drastically changed by the ir - momentum cut . in the ir - momentum gluon cut , the fp spectrum changes from original smooth one to the multi - peaks , which converges into free - field limit .
we also note that the changes of the fp spectrum occurs in the whole eigenvalue region by low - momentum gluon cut .
these results indicate the importance of the ir - momentum gluon for non - trivial structure of the fp eigenmodes .
we comment on change of the fp eigenmodes in the continuum limit . by the ir - momentum cut ,
the fp spectrum becomes the comb - like shapes , since the spectrum converges into the discrete free - field one on lattice .
however , in the continuum limit , the fp spectrum changes into smooth free - field form @xmath127 from near - zero enhanced spectrum in qcd .
therefore , by the ir - momentum cut , the modification of the spectrum would be smaller in the continuum theory than the discrete case as shown in fig.[figfpspectrumir ] . in this paper
, we have investigated the fp eigenmodes in terms of the gluon - momentum components .
it is also interesting to investigate qcd properties in terms of the fp eigenmode , since the fp eigenmodes are considered to be important for confinement in the coulomb gauge .
is supported by a grant - in - aid for jsps fellows [ no .
23 - 752 ] , and h.s .
is supported in part by the grant for scientific research [ ( c ) no.23540306 , priority areas `` new hadrons '' ( e01:21105006 ) ] from the ministry of education , culture , science and technology ( mext ) of japan .
this work is supported by the global coe program , the next generation of physics , spun from universality and emergence " .
the lattice qcd calculations have been done on nec - sx8 and nec - sx9 at osaka university .
h. suganuma , t.t .
takahashi , and h. ichie , _ color confinement and hadrons in quantum chromodynamics _ ( world scientific , singapore , 2004 ) , p.249 ; t.t .
takahashi , h. suganuma , y. nemoto , and h. matsufuru , phys .
d*65 * , 114509 ( 2002 ) ; t.t .
takahashi , h. matsufuru , y. nemoto , and h. suganuma , phys .
lett . * 86*,18 ( 2001 ) . | we analyze the relation between faddeev - popov eigenmodes and gluon - momentum components in the coulomb gauge using su(3 ) lattice qcd . in the coulomb gauge ,
the color - coulomb energy is largely enhanced by near - zero faddeev - popov eigenmodes , which would lead to the confining potential . by the ultraviolet - momentum gluon cut ,
the color - coulomb energy and the faddeev - popov spectrum are almost unchanged .
in contrast to the ultraviolet insensitivity , the color - coulomb energy and the faddeev - popov eigenmodes drastically change by infrared - momentum gluon cut . without infrared gluons , the color - coulomb energy tends to become non - confining , and near - zero faddeev - popov eigenmodes vanish .
we also investigate the full fp eigenmodes , and find that infrared gluons widely influence both high and low faddeev - popov eigenmodes . |
chiral effective field theory ( @xmath3 ) for nuclear physics provides a theoretical framework for a common description of various nuclear processes through the systematic generation of two- three- and many- body interactions and currents .
undoubtedly , it has allowed advances in ab initio structure calculations of light @xcite and medium mass @xcite atomic nuclei .
the quantitative predictions of @xmath3 depend on the numerical values of a set of low - energy constants ( lecs ) , which have become a limiting factor in medium mass nuclei where current sets of lecs fail to simultaneously predict binding , spectra , and radii .
it is therefore relevant to constrain these lecs such that all predictions within the realm of applicability of @xmath3 can be quantified together with statistical uncertainties .
this is important for advancing modern many - body calculations into regions of the nuclear chart and the physical processes where experimental data for verification is limited , such as neutrino - less double beta decay or structure and reactions near the neutron drip line .
we present results that constrain the @xmath2-sector of @xmath3 , with accompanying confidence intervals ( ci ) , up to fourth order in the expansion of the effective lagrangian . at this order ,
the sub - leading long range three - nucleon interaction enters , and will thus be fully constrained by the results presented here .
further , many operator currents , such as the axial - vector current , depend on only @xmath2 lecs and three nucleon contact lecs
. therefore careful uncertainty quantification of the @xmath2 lecs is also critical for comparing processes driven by these currents to experimental cross sections and decay measurements .
the @xmath3 interaction lagrangian @xmath4 for atomic nuclei can be separated into two terms @xmath5 , and the two different contributions each depend explicitly on a distinctive set of lecs .
the first term parametrizes the short - ranged contact - interactions and the second term describes the long - ranged and pion - mediated part of the two- and three - nucleon interaction .
while the @xmath6-contact sector must be constrained using nucleon - nucleon data , the lecs in @xmath7 can be determined from experimental @xmath2 scattering - data , completely separately from the @xmath6 terms .
this is one example of how @xmath3 can link separate physical processes that are relevant for the description of atomic nuclei .
previous constraints for the @xmath2 lecs have been determined from peripheral @xmath6 scattering phase shifts @xcite or @xmath2 elastic scattering phase shifts @xcite .
indeed , these efforts have produced various sets of lecs that closely reproduce the respective phase shift analyses ( either @xmath6 @xcite or @xmath2 @xcite ) ; however , the lack of reliable uncertainties on the input phase shifts prevents meaningful uncertainty quantification of the @xmath2 lecs .
it should be noted that available scattering phase shifts are not experientially measured data , but the result of a partial - wave analysis of measured data .
in contrast with previous determinations of the @xmath2 lecs , the analysis presented here is grounded in experimental scattering data .
this allows us to estimate meaningful statistical uncertainties . for the first time we can therefore explore the consistency of @xmath3 by predicting ci s for the peripheral @xmath6 phase shifts determined by @xmath2-data .
r d,-1 | r d,-1 | r d,-1 & & + & & + @xmath8 & -1.40,0.12 & @xmath9 & + 5.80,0.14 & @xmath10 & + 1.53,0.31 + @xmath11 & + 1.71,0.33 & @xmath12 & -5.66,0.08 & @xmath13 & -11.91,0.87 + @xmath14 & -4.56,0.11 & @xmath15 & + 0.03,0.06 & @xmath16 & + 11.43,1.23 + @xmath17 & + 3.72,0.27 & @xmath18 & -11.50,0.12 & @xmath19 & + 0.73,0.51 + & & & & @xmath20 & + 0.57,1.36 +
we seek a set of @xmath2 lecs @xmath21 that minimize the least - squares objective function : @xmath22 where @xmath23 denotes the value of the scattering observable computed from @xmath3 , while @xmath24 and @xmath25 denotes the experimentally measured value and uncertainty , respectively , for the corresponding observable .
@xmath26 is a vector of lecs spanning all included @xmath27 , @xmath28 , and @xmath29 lecs .
@xmath30 is a vector of normalization coefficients @xmath31 , where all points from a single experimental angular distribution share the same @xmath31 ; @xmath32 encodes uncertainty of the experimental systematics for @xmath31 .
the number of degrees of freedom is given by @xmath33 , where @xmath34 is the number of data included in the fit , @xmath35 is the number of lecs being fit , and @xmath36 is the number of unknown normalization coefficients and @xmath37 is the number of floated coefficients ( contribute no residual term in @xmath38 as @xmath39 ) . for a given experiment @xmath40 , the included data points run over a series of scattering angles at a fixed lab frame momentum @xmath41 . for our fitting dataset
, we adopt the database from the most recent @xmath2 partial wave analysis @xcite , referred to as wi08 . by construction , @xmath3 is a low - energy theory , therefore we exclude data with lab - frame momentum @xmath42 .
this leaves us with an experimental database consisting of differential scattering cross - sections and polarization cross sections from @xmath43 and @xmath44 processes . in total
, there are @xmath45 data points , consisting of @xmath46 differential unpolarized cross - sections and @xmath47 differential singly - polarized cross - sections .
there are @xmath48 normalization coefficients , with @xmath49 floated coefficients .
there are other measurable observables , such as the spin rotation parameters , but experimental data only exists for momentum well beyond the range of validity of the eft .
the cutoff in lab frame momentum ( @xmath50 ) was chosen such that increasing or decreasing the cutoff would lead to a larger minimum value of @xmath51 , maximizing amount of included data while avoiding fitting past the radius of convergence of the eft .
for the calculated observables , we use the strong amplitudes presented in ref .
@xcite ( refs .
@xcite give a more complete presentation of the @xmath2 scattering amplitudes , but use a different power counting scheme for relativistic corrections . ) for the strong amplitude , we adopt their conventions for fixing @xmath52 , absorbing @xmath53 , @xmath54 into @xmath55 .
we also work with exact isospin symmetry and use an averaged pion mass ( @xmath56 ) .
we adopt the electromagnetic treatment that is used in the wi08 partial wave analysis , which is described in detail inrefs . @xcite .
for the electromagnetic corrections , we explicitly break isospin symmetry and use the physical pion masses within the coulomb amplitudes .
actual fits were performed using the tao package @xcite .
the central values and @xmath57 uncertainties of the @xmath2 lecs up to fourth order in @xmath3 from fitting against scattering data are presented in table [ tab : lecs ] . for this fit
, we find that @xmath58 . as a comparison , the lecs from the fit against wi08 phase shifts of ref .
@xcite generate @xmath59 with respect to our objective function .
not surprisingly , our fit will reproduce experimental data better than the fits with respect to phase shifts . while our lecc are consistent with the spread of previous analyses , we find that no single analysis is entirely consistent with our fit at the @xmath60 confidence level , though the wi08 @xmath2 phase shift fit from ref .
@xcite lies just outside our interval .
c| d .. -1 d .. -1 d .. -1 d .. -1 d .. -1 d .. -1 d .. -1 d .. -1 d .. -1 d .. -1 d .. -1 d .. -1 d .. -1 lec & & & & & & & & & & & & & + @xmath8 & + 0.03 & & + 0.24 & + 0.56 & + 0.51 & -0.33 & -0.48 & -0.45 & -0.31 & + 0.37 & -0.85 & + 0.15 & -0.54 + @xmath11 & + 0.07 & + 0.21 & & + 0.55 & + 0.54 & -0.37 & -0.48 & -0.56 & -0.32 & + 0.56 & -0.96 & + 0.18 & -0.54 + @xmath14 & + 0.01 & -0.01 & + 0.02 & & -0.02 & + 0.12 & -0.06 & + 0.26 & + 0.08 & -0.58 & + 0.26 & -0.12 & -0.08 + @xmath17 & + 0.03 & + 0.09 & + 0.01 & + 0.14 & & -0.75 & -0.77 & -0.74 & -0.39 & + 0.36 & -0.55 & + 0.18 & -0.94 + @xmath9 & + 0.01 & + 0.05 & -0.00 & + 0.07 & + 0.03 & & -0.76 & -0.75 & -0.43 & + 0.46 & -0.57 & + 0.18 & -0.91 + @xmath12 & -0.01 & -0.02 & + 0.00 & -0.03 & -0.02 & + 0.01 & & + 0.67 & + 0.43 & -0.45 & + 0.44 & -0.24 & + 0.74 + @xmath15 & -0.01 & -0.02 & -0.00 & -0.02 & -0.01 & + 0.00 & + 0.00 & & + 0.25 & -0.28 & + 0.48 & -0.04 & + 0.69 + @xmath18 & -0.01 & -0.04 & + 0.01 & -0.04 & -0.02 & + 0.01 & + 0.01 & + 0.02 & & -0.52 & + 0.62 & -0.47 & + 0.81 + @xmath10 & -0.02 & -0.05 & + 0.00 & -0.05 & -0.03 & + 0.02 & + 0.01 & + 0.02 & + 0.13 & & + 0.50 & -0.28 & + 0.41 + @xmath13 & + 0.07 & + 0.29 & -0.10 & + 0.15 & + 0.10 & -0.05 & -0.02 & -0.09 & -0.33 & + 1.30 & & + 0.32 & -0.40 + @xmath16 & -0.22 & -0.71 & + 0.06 & -0.33 & -0.17 & + 0.07 & + 0.05 & + 0.15 & + 0.30 & -1.40 & + 2.59 & & + 0.56 + @xmath19 & + 0.01 & + 0.05 & -0.01 & + 0.04 & + 0.02 & -0.02 & -0.00 & -0.04 & -0.06 & + 0.22 & -0.24 & + 0.36 & + @xmath20 & -0.14 & -0.41 & -0.02 & -0.58 & -0.27 & + 0.13 & + 0.08 & + 0.21 & + 0.24 & -0.75 & + 1.48 & -0.49 & + 2.70 for uncertainty analysis , we apply a standard gradient expansion of @xmath61 at the optimum @xmath21(see ref .
@xcite and references therein for further detail ) : @xmath62 @xmath63 where @xmath61 is the full ( not reduced by @xmath64 ) objective function .
if the residual vectors ( @xmath65 and @xmath38 ) are normally distributed , then the covariance matrix of our fit is given by @xmath66 .
the covariance matrix ( @xmath67 ) and correlation matrix @xmath68 are presented in table [ tab : covcor ] .
+ figure [ fig : pin : obs ] shows calculations of @xmath2 scattering observables using our lecs fit data as well as lecs from the phase shift fit of ref .
@xcite . at smaller @xmath69 ,
the difference is minimal , but it is clear at higher momentum that phase shifts fits are inadequate .
this is especially apparent in the calculations of the polarization ( p ) and spin rotation parameters ( a and b ) , suggesting a phase shift fit may not adequately capture the underlying tensor effects that are critical to nuclear observables . as a validation of our analysis ,
we plot @xmath2 partial wave phase shifts in fig .
[ fig : pin : phaseshifts ] and compare with two different partial wave analyses(wi08 @xcite and ka84 @xcite ) the blue bands denotes the @xmath60 ci from our fit . for s- and p - waves
, we have good agreement with the wi08 partial wave analysis . in the d - waves , where the @xmath70 lecs have significant contributions
, we find poor agreement in the @xmath71 channels ( d13 and d33 ) .
to better understand where this error in the @xmath71 d - waves is coming from , we can use the eigenstate of the covariance matrix to gain insight on the quality of the lec constraints : @xmath72 the vectors of lec combinations , @xmath73 , form pseudo lecs that are completely uncorrelated , and eigenvalues @xmath74 are effective variances of these pseudo lecs . in table
[ tab : pca ] , we examine the lec content of the least constrained eigenvectors ( those with largest eigenvalue ) .
it is clear that the largest eigenvectors are dominated by the @xmath75 lecs , and in particular the pair @xmath16 , @xmath20 which dominate the two most significant vectors .
this indicates that @xmath16 and @xmath20 are poorly constrained . within the scattering amplitudes ,
@xmath16 and @xmath20 span the non - spin - flip and spin - flip amplitudes respectively @xcite , suggesting that a lack of low momentum data for spin observables could be at fault .
.[tab : pca ] eigenvalues ( @xmath74 ) of the covariance matrix , with contributions to eigenvector ( @xmath76 ) summed over lecs of same chiral order .
these eigenvalue correspond to the variances @xmath77 of effective uncorrelated pseudo lecs . for this analysis the covariance matrix was reduced to a dimensionless form using appropriate powers of the nucleon mass .
[ cols= " > , > , > , > , > , > , > " , ]
accurate calculations accompanied with quantitative uncertainty estimates is of - course one of the principal goals of low - energy nuclear theory .
this is critical to the study of nuclei near the neutron drip - line or for @xmath78 decay where this insufficient data to validate many - body calculations . as a first step towards quantifying the accuracy of the @xmath2 sector of the @xmath6 interaction
, we will predict the confidence intervals of the peripheral ( @xmath79 ) @xmath6 scattering phase shifts at @xmath80 .
these partial waves are fully determined by the long - ranged pion physics .
it should also be noted that this is the first @xmath80 calculation of these phase shifts that is actually grounded in experimental scattering data .
figure [ fig : uqnn ] shows the predicted @xmath60 cis of the proton- neutron elastic scattering phase shifts for all total angular momentum @xmath81 partial waves .
for comparison , phase shifts from two different partial wave analyses are presented , pwa93 @xcite and sp07 @xcite .
we also compare to phase shifts computed using the high - precision idaho-@xmath80 interaction @xcite , which reproduces the sm99 @xcite @xmath6 database with @xmath82 , and to @xmath80 @xmath6 interactions computed using the lecs from ref .
our lecs perform quite favorably compared the @xmath2 phase shift fit , and surprisingly favorable in many channels to the phase shifts computed from idaho-@xmath80 . for many channels ,
our error bands are remarkably small .
we constrained the @xmath2 lecs using experimental data with @xmath83 , generating a set of lecs with error estimates that are compatible for use with modern @xmath803 hamiltonians and currents .
we find that our lower order lecs are of natural size , while the higher order lecs tend to be unnaturally large ( or small in the case of @xmath15 . )
we find that even with fairly large error bars for some lecs , the proton - neutron peripheral phase shifts are very well constrained at lab scattering energies below @xmath84 .
the @xmath2 lecs not only fix the long range part of the @xmath80 three body force , but our analysis provides a means to examine uncertainty of the three body force in nuclear bound states as well as uncertainty from currents in observables . progressing forward , simultaneous constraints of @xmath2 and @xmath6 lecs with quantitative statical analysis will yield predictions of few- and many - body systems with quantified uncertainty and is a topic for many future investigations .
the authors would like to thank h. krebs , t. papenbrock , d. phillips , and r. workman for their useful discussion .
this work was supported in part by the u.s .
department of energy ( doe ) under grant nos .
defg02 - 96er40963 ( university of tennessee ) , de - sc0008499 ( nuclei scidac collaboration ) , oak ridge national laboratory the research council of norway under contract isp - fysikk/216699 , and by the european research council under the european community s seventh framework programme ( fp7/2007 - 2013 ) erc grant agreement no . 240603 .
oak ridge national laboratory is supported by the doe office of science under contract no .
de- ac05 - 00or22725 .
28ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ] + 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1103/physrevc.87.014327 [ * * , ( ) ] , link:\doibase 10.1103/physrevlett.109.052501 [ * * , ( ) ] link:\doibase 10.1103/physrevc.87.034307 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.032502 [ * * , ( ) ] , link:\doibase 10.1103/physrevc.66.014002 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.110.192502 [ * * , ( ) ] link:\doibase 10.1016/s0375 - 9474(99)00813 - 1 [ * * , ( ) ] link:\doibase 10.1016/s0375 - 9474(00)00199 - 8 [ * * , ( ) ] link:\doibase 10.1103/physrevc.85.054006 [ * * , ( ) ] link:\doibase 10.1103/physrevc.48.792 [ * * , ( ) ] link:\doibase 10.1103/physrevc.76.025209 [ * * , ( ) ] link:\doibase 10.1007/bf01560295 [ * * , ( ) ] link:\doibase 10.1016/0375 - 9474(86)90438 - 0 [ * * , ( ) ] link:\doibase 10.1103/physrevc.86.035202 [ * * , ( ) ] @noop link:\doibase 10.1016/s0375 - 9474(98)00452 - 7 [ * * , ( ) ] link:\doibase 10.1006/aphy.2000.6059 [ * * , ( ) ] link:\doibase 10.1016/s0375 - 9474(01)00977 - 0 [ * * , ( ) ] link:\doibase 10.5169/seals-114961 [ * * ( ) , 10.5169/seals-114961 ] link:\doibase 10.1016/0550 - 3213(74)90538 - 0 [ * * , ( ) ] link:\doibase 10.1103/physrevd.15.725 [ * * , ( ) ] link:\doibase 10.1016/0550 - 3213(73)90591 - 9 [ * * , ( ) ] @noop _ _ , ( , ) link:\doibase 10.1088/0954 - 3899/41/7/074001 [ * * , ( ) ] , link:\doibase 10.1007/b19946 [ _ _ ] , edited by , , vol .
( , , ) link:\doibase 10.1103/physrevc.63.024001 [ * * , ( ) ] link:\doibase 10.1103/physrevc.68.041001 [ * * ( ) , 10.1103/physrevc.68.041001 ] | we extract the statistical uncertainties for the pion - nucleon ( @xmath0 ) low energy constants ( lecs ) up to fourth order @xmath1 in the chiral expansion of the nuclear effective lagrangian .
the lecs are optimized with respect to experimental scattering data . for comparison
, we also present an uncertainty quantification that is based solely on @xmath2 scattering phase shifts .
statistical errors on the lecs are critical in order to estimate the subsequent uncertainties in _ ab initio _ modeling of light and medium mass nuclei which exploit chiral effective field theory . as an example of the this , we present the first complete predictions with uncertainty quantification of peripheral phase shifts of elastic proton - neutron scattering . |
in 1966 , forney introduced the concept of concatenated codes @xcite .
it was generalized in 1976 by blokh and zyablov to generalized concatenated ( gc ) codes @xcite .
the gc approach allows to design powerful codes with large block lengths using short , and thus easily decodable , component codes . the designed distance and the performance of gc codes
can be easily estimated theoretically .
this allows to design gc codes for applications like e.g. optical lines , where block error rates in the order of @xmath2 are required , a region , where simulations are not feasible .
gc codes can be decoded up to half their minimum distance using a sufficiently large number of decoding attempts for each outer code in the blokh
zyablov dumer algorithm ( bzda ) @xcite ,
see also dumer @xcite .
we should also mention papers by sorger @xcite , and ktter @xcite who suggested interesting modifications of a bmd decoder in such a way that multi - attempt decoding of the outer code can be made in one step .
nielsen suggested in @xcite to use the guruswami
sudan list decoding algorithm @xcite for decoding the outer codes and has shown that in this case also one decoding attempt is sufficient to allow decoding up to half the minimum distance of the gc code . in this paper
, we employ another idea , which allows to decrease the number of outer decodings , but also to skip many decodings of the inner code .
this idea is based on interleaved reed solomon ( irs ) codes @xcite .
other aspects of using irs codes in concatenated codes were considered in @xcite .
the rest of our paper is organized as follows : in section [ section : gcc ] we explain gc codes as well as the required notations and assumptions . section [ section : irs ] explains irs codes and where they appear within gc codes .
section [ section : bzouterrs ] gives an overview of the bzda as introduced in @xcite .
a generalization of this is given in section [ section : bzouterirs ] , leading to a new algorithm which maintains the error - correcting performance of gc codes while reducing the number of outer decodings and skipping many inner decodings . in section [ section :
example ] we illustrate our results by means of some examples .
encoding of a gc code @xmath3 of order @xmath4 is as follows , where we restrict ourselves to outer rs codes @xmath5 , @xmath6 , over the binary extension field @xmath7 and an inner binary block code @xmath8 with dimension @xmath9 .
the first step is outer encoding . for this
we take @xmath4 codewords @xmath10 of the outer codes @xmath11 and put them as rows into an @xmath12 matrix @xmath13 over @xmath7 .
the second step is inner encoding , where the binary counterparts of the columns @xmath14 of @xmath15 are encoded by the inner code to obtain code word columns @xmath16 .
the result of this procedure is a binary @xmath17 matrix @xmath18 , which in turn is a code word of the gc code @xmath19 .
the inner code @xmath20 has the following nested structure : the code words are obtained by encoding an arbitrary binary information vector @xmath21 .
if we fix @xmath22 groups of @xmath23 information bits starting from @xmath24 and encode , we obtain a subcode @xmath25 with distance @xmath26 .
note that @xmath27 due to the special choice of encoders for the subcodes @xmath28 .
obviously , @xmath29
. the minimum distance of the gc code @xmath19 is then lower bounded by its designed distance @xmath30 the channel adds a binary error matrix @xmath31 of weight @xmath32 to the transmitted concatenated code matrix @xmath33 , resulting in a received matrix @xmath34 at the receiver . as output of gc decoding with the bzda we want to obtain a matrix @xmath35 over @xmath7 , which is an estimate of the matrix @xmath15 .
decoding consists of @xmath4 iterations , where iteration @xmath36 is as follows : a. decoding of all columns @xmath37 , @xmath38 , of the received matrix @xmath39 by bmd decoders for the nested subcodes @xmath40 , correcting up to @xmath41 errors and altogether yielding a new estimate @xmath42 for the @xmath43-th row of @xmath44 .
b. execution of @xmath45 attempts of decoding @xmath42 with @xmath46 and a different number of erased symbols in each attempt , which yields a candidate list of at most size @xmath45 . finally , the best
candidate @xmath47 from this list has to be selected using some criterion and inserted into @xmath44 . note that because of this recurrent structure it is sufficient to consider the bzda only for ordinary concatenated codes , as decoding of gc codes simply means repeated application of this special case for the sequence @xmath48 of the nested inner subcodes , @xmath14 , and the corresponding outer codes @xcite .
the details of the bzda for ordinary concatenated codes are described in section [ section : bzouterrs ] . in @xcite
it is shown that if @xmath49 for even @xmath50 , the bzda can decode up to @xmath51 errors in the @xmath43-th iteration and thus by ( [ eqn : designeddistance ] ) also up to @xmath52 errors in a gc code word .
observe that the matrix @xmath15 is just an interleaved set of @xmath4 different rs codes , hence an irs code .
for irs codes an efficient decoding algorithm was suggested in @xcite , which has only @xmath4 times the complexity of the berlekamp
massey algorithm for decoding one single rs code .
the algorithm allows to correct at most @xmath53 erroneous columns of @xmath15 , where @xmath54 is the average minimum distance of the interleaved set of rs codes and @xmath55 the irs decoding algorithm from @xcite yields a decoding failure with some probability .
however , this probability can be made small and is neglected here .
if the complete matrix @xmath15 does not satisfy ( [ eqn : l1 ] ) , we can split it into a number of submatrices with the same length as @xmath15 , which all fulfill ( [ eqn : l1 ] ) and thus can be decoded by the irs decoding algorithm from @xcite .
assume that @xmath56 is such a submatrix of @xmath15 and forms an irs code with average minimum distance @xmath57 , that satisfies both constraint ( [ eqn : l1 ] ) and @xmath58 .
the main idea of applying the irs decoding algorithm from @xcite to gc codes is as follows : we can replace @xmath59 iterations @xmath60 of the bzda by the following single iteration : a. decoding of all columns @xmath37 , @xmath38 , of the received matrix @xmath39 by bmd decoders for the subcodes @xmath61 , correcting up to @xmath62 errors and yielding an estimate @xmath63 for the submatrix @xmath64 .
b. execution of @xmath65 attempts of irs decoding @xmath63 with a different number of erased columns in each attempt , which yields a candidate list of at most size @xmath65 .
finally , the best
candidate @xmath66 from this list has to be selected using some criterion and inserted into @xmath44 . as a result of this method , we skipped @xmath67 inner decodings and
we will show that the number of required decoding attempts for the outer code to guarantee decoding up to @xmath68 channel errors is much smaller than in the original bzda , which in practice means @xmath69 .
eventually , our modified algorithm corrects up to half the minimum distance of the gc code .
in this section , we consider decoding of a _ simple _ concatenated code @xmath19 , which consists of the outer rs code @xmath70 and the inner binary code @xmath20 .
this corresponds to the iteration of the refined bzda from @xcite for a gc code where it holds w.l.o.g @xmath71 , @xmath72 and @xmath73 .
first ( step ( i ) of the bzda ) , we decode the columns @xmath37 of the received matrix @xmath39 by a bmd decoder for @xmath20 , correcting up to @xmath74 errors and yielding code word estimates @xmath75 or decoding failures .
decoding of the outer code @xmath70 is performed with respect to the ordered set of thresholds @xmath76 with @xmath77 . for each decoding attempt
@xmath78 the decoding results of the inner decoder depend on the threshold @xmath79 in the following manner : the symbols @xmath80 delivered to the outer decoder at position @xmath81 are @xmath82 where @xmath37 is the received word in the @xmath81-th column , @xmath75 is the result of inner decoding , @xmath83 maps code words of @xmath20 to the corresponding @xmath84-ary information symbols , and @xmath85 is the symbol for an erasure . as result from outer decoding @xmath86 we obtain the outer code word estimate @xmath87 . from ( [ eqn : rtilde ] ) follows that thresholds with equal integers parts yield equal decoding attempts , so the number of @xmath88 actual attempts may be smaller than the number @xmath89 of thresholds , i.e. @xmath90 . the numbers of decoding errors and erasures occurring at decoding @xmath91 are denoted by @xmath92 and @xmath93 , respectively .
the outer rs code @xmath70 can successfully decode as long as @xmath94 , since we assume outer bmd decoding in this section . for a fixed number @xmath89 of thresholds , the following theorem fixes the optimum values of the thresholds such that the decoding bound of the bzda is maximized .
[ theorem : thresholdsind ] for a concatenated code with outer bmd - decoded rs code and inner bmd - decoded code @xmath8 , the set of thresholds @xmath76 which maximizes the decoding bound is determined by @xmath95 @xmath96 .
if the thresholds are chosen according to ( [ eqn : thresholdsind ] ) , the decoding bound is given by the following theorem in a sense that the transmitted code word is among the elements of the result list @xmath97 .
[ theorem : boundind ] for a concatenated code with outer bmd - decoded rs code @xmath98 and inner bmd - decoded code @xmath8 , the decoding bound is @xmath99 in figure [ fig : outer_irs_even ] , the decoding bound ( [ eqn : boundind ] ) is plotted with circles versus the number of thresholds @xmath89 for an example with @xmath100 , @xmath101 .
it can clearly be seen that the bound reaches @xmath102 , i.e. half the minimum distance of the concatenated code with increasing number of thresholds .
the bound obviously only depends on the greatest threshold @xmath103 .
if we hypothesize that the number of thresholds tends to infinity , we can see that for the greatest threshold @xmath104 but as @xmath105 , we know that the greatest possible integer threshold @xmath106 is @xmath107 if @xmath50 is even , and @xmath108 if @xmath50 is odd .
this allows to state the following theorem , which confirms our observation from figure [ fig : outer_irs_even ] .
[ theorem : maxboundind ] if the number of thresholds @xmath89 tends to infinity , the decoding bound of the bzda for a concatenated code @xmath19 with outer bmd - decoded rs code @xmath98 and inner bmd - decoded code @xmath8 is @xmath109 the decoding bound ( [ eqn : boundind ] ) is non - decreasing in @xmath89 , hence it assumes its maximum at @xmath110 .
consider two cases : a. @xmath50 is even , thus the greatest possible integer threshold is @xmath107 and @xmath111 .
b. @xmath50 is odd , hence the greatest possible integer threshold is @xmath108 and @xmath112 . in the following ,
we restrict ourselves to binary error matrices @xmath31 meeting ( [ eqn : maxboundind ] ) . to obtain decoding bound ( [ eqn : maxboundind ] ) , the greatest possible integer threshold needs to be among the threshold set . for even @xmath50
this greatest integer threshold is @xmath113 , which is strictly smaller than the limit @xmath114 . by the following lemma
it can be reached already for a rather small value of @xmath89 .
[ lemma : indeven ] for a concatenated code with inner bmd - decoded code @xmath8 and outer bmd - decoded rs code @xmath70 the greatest possible integer threshold @xmath115 is reached if @xmath116 .
solve @xmath117 for @xmath89 .
we can thus obtain ( [ eqn : maxboundind ] ) with only @xmath118 thresholds according to ( [ eqn : thresholdsind ] ) if @xmath50 is even .
if however @xmath50 is odd , the greatest possible integer threshold is @xmath119 , i.e. the limit @xmath114 itself .
it can obviously only be reached for an infinte number of thresholds .
but the number of integers below @xmath120 is @xmath108 , hence the number of actual decoding attempts is upper bounded by @xmath108 .
it follows that even though in the @xmath50 odd case the number of required thresholds is infinite , only @xmath121 outer decoding attempts are sufficient to achieve decoding bound ( [ eqn : maxboundind ] ) . up to now , we only know that the transmitted outer code word @xmath122 is _ somewhere _ within the result list @xmath123 of the bzda if ( [ eqn : maxboundind ] ) is fulfilled . the following lemma provides a means of exactly determining its position among the elements of @xmath123 .
[ lemma : selection ] let @xmath124 with @xmath125 and @xmath126 .
assume @xmath127 and that @xmath128 is a threshold with @xmath129 .
then @xmath130 and @xmath131 the lemma guarantees that only the transmitted outer code word @xmath132 fulfills ( [ eqn : selection ] ) , i.e. that no further decoding attempts have to be executed as soon as ( [ eqn : selection ] ) is fulfilled for the smallest threshold index @xmath96 .
then , we set @xmath133 and choose @xmath134 .
now we consider the case where @xmath70 is an irs code , i.e. a row - wise arrangement of @xmath135 rs codes of equal length but potentially different dimensions , which are decoded collaboratively as described in section [ section : irs ] .
this allows @xmath70 to correct a larger number of errors leading to a decoding success whilst @xmath136 , where @xmath137 .
this means bounded distance ( bd ) decoding .
our aim now is to derive formulae corresponding to ( [ eqn : thresholdsind ] ) and ( [ eqn : boundind ] ) for this specific case . in doing so
, we generalize the approach for outer bmd decoding from @xcite .
the procedure is as follows : let @xmath138 be the smallest number of channel errors for a given set of thresholds @xmath139 , such that all decoding attempts @xmath140 fail , i.e. such that @xmath141 we determine @xmath142 under the condition that ( [ eqn : errorforall ] ) is fulfilled .
then , we find the set @xmath139 of thresholds which maximizes this minimum , i.e. the set of thresholds , which maximizes the decoding bound .
this set is determined by the expression @xmath143 the detailed derivation is too involved to be presented here , so we confine ourselves to the results in form of the following theorems . [
theorem : thresholds ] for a concatenated code @xmath19 with outer collaboratively decoded irs code @xmath70 consisting of @xmath4 rs codes and inner bmd - decoded code @xmath8 , the set of thresholds @xmath76 which maximizes the decoding bound is defined by @xmath144 with @xmath145 @xmath140 , where @xmath89 is the number of thresholds and @xmath146 .
[ theorem : boundcol ] for a concatenated code @xmath19 with outer collaboratively decoded irs code @xmath98 consisting of @xmath4 rs codes and @xmath89 thresholds chosen as in ( [ eqn : thresholdscol ] ) , the decoding bound is given by @xmath147 by theorem [ theorem : boundcol ] the decoding bound only depends on threshold @xmath103 , the greatest one among the ordered threshold set @xmath76 . hence to maximize the decoding bound ( [ eqn : boundcol ] ) we have to maximize @xmath103 . since the threshold location function ( [ eqn : thresholdscol ] )
is non - decreasing , the greatest threshold occurs for @xmath110 , and is @xmath148 .
the following theorem states the decoding bound for this greatest possible threshold .
[ theorem : irshd ] let @xmath19 be a concatenated code with inner bmd - decoded code @xmath8 and outer irs code @xmath70 with @xmath149 . if the maximum possible integer threshold is among the threshold set , the decoding bound is given by @xmath150 inserting the the integer parts @xmath151 and @xmath152 , respectively , of the greatest possible thresholds into bound ( [ eqn : boundcol ] ) proves the statement . for even @xmath50
, the greatest possible integer threshold @xmath153 already is reached considering a finite number of thresholds , i.e. if @xmath154 .
thus for even @xmath50 the finite threshold set @xmath155 is sufficient to obtain the maximum of ( [ eqn : boundcol ] ) . if on the other hand @xmath50 is odd , the greatest possible integer threshold @xmath156 is @xmath114 itself , hence the number of required thresholds in fact is infinite .
but since we know that decoding attempts corresponding to thresholds with equal integer parts coincide , we can skip all thresholds within the interval @xmath157 by the following lemma .
[ lemma : neededthresholds ] for a concatenated code with inner bmd - decoded code @xmath8 and and outer irs code @xmath70 with @xmath4 collaboratively decoded rs codes @xmath158 is reached if @xmath159 .
if @xmath160 , then the threshold location function ( [ eqn : thresholdscol ] ) becomes @xmath161 .
but @xmath162 . by lemma [ lemma : neededthresholds ]
we know that all thresholds @xmath163 in the range @xmath164 have equal integer parts and therefore can be omitted .
thus , instead of the infinite threshold set @xmath165 it is equivalent to consider the finite set @xmath166 with only @xmath167 elements . we know
that if we utilize the sets @xmath168 and @xmath169 of thresholds according to ( [ eqn : thresholdscol ] ) for even and odd inner minimum distance @xmath50 , respectively , we can decode up to half the minimum distance of the concatenated code @xmath19 .
however , the integer parts not of all the thresholds among the sets are necessarily pairwise different . since decoding attempts in respect to thresholds with equal integer parts coincide , the number @xmath88 of actual decoding attempts which need to be executed to decode up to half the minimum distance of @xmath19 may be smaller than the number of thresholds .
we can calculate it explicitly by @xmath170
our results are subsumed using the following examples . we assume a concatenated code @xmath19 consisting of an inner code @xmath8 and an outer code @xmath98 consisting of @xmath4 rows containing code words of the rs code @xmath171 . for _
even _ inner minimum distance @xmath101 the decoding bounds ( [ eqn : boundind ] ) for independent outer decoding and ( [ eqn : boundcol ] ) for collaborative outer decoding , respectively , depending on the number @xmath89 of thresholds are shown in figure [ fig : outer_irs_even ] . according to lemma [ lemma : indeven ] , for independent outer decoding @xmath172 thresholds are sufficient to decode up to half the minimum distance of @xmath19 .
if collaborative decoding of @xmath173 outer rs codes is applied , we can calculate the number of required thresholds by @xmath174 and get @xmath175 .
both values are confirmed by the bounds in figure [ fig : outer_irs_even ] .
if the rs codes are outer codes @xmath176 of a gc code as described in section [ section : gcc ] , which fulfill ( [ eqn : l1 ] ) , the saving in terms of operations is even greater .
besides the @xmath177 saved outer decoding attempts , the number of inner decodings can then be cut down by @xmath178 . note that decoding one irs code with @xmath4 interleaved rs codes with the algorithm from @xcite has the same complexity as decoding the @xmath4 rs codes independently .
thus , our comparison of both constructions is fair in terms of complexity .
figure [ fig : attempts ] shows the number of actual decoding attempts @xmath88 as well as the number @xmath180 of thresholds for some reasonable _ odd _ inner minimum distances @xmath50 .
collaborative decoding of @xmath173 and @xmath181 outer rs codes is considered . for independent outer decoding
as described in section [ section : bzouterrs ] , @xmath88 grows linearly with @xmath50 .
it diminishes to at most @xmath182 already for an outer irs code with @xmath173 . for an outer irs code with @xmath181 already @xmath183 decoding attempts are sufficient to decode up to half the minimum distance of @xmath19 over the full range of all considered odd inner minimum distances @xmath184 $ ] .
g. schmidt , v. r. sidorenko , and m. bossert , `` interleaved reed
solomon codes in concatenated code designs , '' in _ proc .
ieee itsoc inform .
theory workshop _ , ( rotorua , new zealand ) , pp . 187191 , august 2005 .
g. schmidt , v. r. sidorenko , and m. bossert , `` collaborative decoding of interleaved reed solomon codes and concatenated code designs . ''
preprint , available online at arxiv , arxiv : cs.it/0610074 , 2006 . | generalized concatenated codes are a code construction consisting of a number of outer codes whose code symbols are protected by an inner code . as outer codes , we assume the most frequently used reed solomon codes ; as inner code , we assume some linear block code which can be decoded up to half its minimum distance . decoding up to half the minimum distance of generalized concatenated codes
is classically achieved by the blokh zyablov
dumer algorithm , which iteratively decodes by first using the inner decoder to get an estimate of the outer code words and then using an outer error / erasure decoder with a varying number of erasures determined by a set of pre - calculated thresholds . in this paper , a modified version of the blokh
zyablov
dumer algorithm is proposed , which exploits the fact that a number of outer reed
solomon codes with average minimum distance @xmath0 can be grouped into one single interleaved reed
solomon code which can be decoded beyond @xmath1 .
this allows to skip a number of decoding iterations on the one hand and to reduce the complexity of each decoding iteration significantly while maintaining the decoding performance on the other . |
in regression learning problems , we are given data @xmath13 in @xmath14 where @xmath15 is a bounded subset of @xmath16 and @xmath17 is a bounded subset of @xmath18 .
we assume this data is chosen iid ( independently and identically distributed ) according to an unknown probability distribution @xmath19 .
we say that @xmath10 is a ` position ' , and @xmath20 is a ` label ' .
these data points may be , for example , images of people s faces in pixel space with a person s age as the corresponding label , or auto - regressive time series data ( @xcite , @xcite ) .
the output of a learning algorithm is a decision function @xmath21 .
even though we only know @xmath5 data points from distribution @xmath19 , we hope to construct @xmath22 which will be able to _ generalize _ to unobserved points in the distribution .
this means we would like @xmath22 to predict the value of @xmath20 for any given value of @xmath23 .
since we want our function @xmath22 to fit the data accurately and also have this generalization ability , we refer to vapnik s structural risk minimization ( srm ) principle ( @xcite , @xcite ) . in srm , we limit our choice of functions @xmath22 so they are chosen from a class @xmath24 , of finite capacity ( i.e. finite vc dimension ) .
otherwise , we can not hope to choose a function @xmath22 which has generalization ability - we would overfit the data .
one convenient way to implement srm is to let @xmath24 be a ball within a reproducing kernel hilbert space ( rkhs ) @xmath25 , with norm @xmath26 . in this form
, we have an ivanov regularization problem ; one can show that the solution is always the minimizer of a corresponding tikhonov regularization problem .
algorithms for classification and regression solve this tikhonov regularization problem , so that the decision function is given by @xmath27 ( @xcite , @xcite ) , where @xmath28 @xmath12 is called the regularized risk functional .
note that we define our rkhs @xmath25 so that @xmath29 iff @xmath30 is finite .
thus , minimizing over @xmath29 is equivalent to minimizing over all functions @xmath22 . the first term in @xmath12
is called the empirical risk , and @xmath31 is a pre - determined loss function .
we will generally use the least squares loss function @xmath32 , but a similar analysis can be performed for other loss functions . the second term is called the regularization term , and @xmath33 is called the regularization parameter ; one always takes @xmath34 . here
, @xmath33 can be viewed as the trade - off between accuracy and generalization .
if @xmath33 is very small , we are minimizing the empirical risk , increasing the accuracy of our model to the data , and possibly overfitting .
if @xmath33 is very large , our algorithm will generalize at the expense of accuracy . in a sense , @xmath33 controls the capacity of the function class from which @xmath22 is chosen : the larger @xmath33 is , the smaller the radius of the ball @xmath24 in @xmath25 . in practice ,
@xmath33 is often chosen empirically , perhaps to minimize the leave - one - out error on a training set .
another interpretation of this functional is through the eyes of algorithmic stability , as described by bousquet and elisseeff ( @xcite ) . here
, the regularization term prevents the algorithm from being sensitive to the replacement of one data point . in either case ,
the regularization problem is well - posed only when @xmath33 is strictly greater than 0 .
we assume that the labels @xmath20 are ` noisy ' , in the sense that there is a marginal distribution @xmath35 for each @xmath10 .
we denote the expectation value of the label y for position x as @xmath36 , and we denote the marginal distribution along the @xmath10-axis as @xmath37 .
( this is the distribution of @xmath19 after integrating over the @xmath20 values . ) for the case when @xmath38 , we show convergence of @xmath27 to a function @xmath8 as the regularization term vanishes , provided @xmath39 ; i.e. we need to find conditions on the simultaneous convergence @xmath38 and @xmath40 so that @xmath41 . here , the function @xmath8 is defined by : @xmath42 ^
2|\mthx\right ) d\mu(\mthx).\end{aligned}\ ] ] in other words , @xmath43 is the minimizer of the _ ` actual risk'_. since we are using the least squares loss function , this minimizer is simply the expectation of @xmath20 for each @xmath10 ; @xmath44 .
we assume that we have chosen a rkhs which is large enough to contain @xmath45 .
in other words , @xmath46 . this is not an exceedingly strong assumption ; in fact , many popular kernels ( e.g. gaussian kernels ) can produce rkhs of arbitrarily high vc dimension .
although @xmath45 may not be in @xmath25 for every case , @xmath45 will be in @xmath25 for most smooth processes which have bounded noise , as long as we implement a sufficiently powerful rkhs .
for the fixed @xmath5 case , we may express label @xmath20 for position @xmath10 as the random variable @xmath47 , where @xmath48 is a deterministic function assumed to be in @xmath25 , and @xmath49 is random noise with some probability distribution , with @xmath49 and @xmath50 independent if @xmath51 .
we denote the vector of noise values as @xmath52 . in order to force the noise to vanish
, we will assume the noise is generated by a fixed random process generating noise with norm bounded by @xmath53 almost surely , and we will only shrink its amplitude .
since the noise is generated by this fixed process , the theorem will hold whenever the noise is bounded , and thus , if the noise is bounded almost surely , the theorem will hold almost surely . using
the least squared loss and making the noise explicit , our algorithm becomes : @xmath54 for @xmath34 , the minimizer of @xmath12 is unique , because @xmath55 is strictly convex .
since the noise is random , @xmath27 is still a random variable .
our goal is to show ` stability ' for this algorithm , i.e. we need to find a set of conditions on the simultaneous convergence @xmath56 which allows @xmath57 , where @xmath4 is the element of @xmath25 with minimal norm that has zero empirical risk when noise is not present .
( since we assume that @xmath58 , @xmath59 itself minimizes @xmath12 when @xmath60 . since many functions in @xmath25 vanish at all the @xmath61 , there may be infinitely many functions with zero empirical risk ; our algorithm will converge to the one with the smallest rkhs norm . )
intuitively , this stability analysis demonstrates that there s no inherent error in our algorithm when noise or regularization is present , and that a small amount of noise or regularization can not dramatically disrupt the algorithm s output .
this type of stability is different from the ` algorithmic stability ' of devroye(@xcite ) .
algorithmic stability measures the variability of an algorithm s output as the data set changes .
our type of stability determines whether the algorithm s output changes dramatically when noise or regularization is present .
algorithmic stability is a property of one particular algorithm for one particular distribution of data .
our stability is not - the distribution changes as noise is removed , and the algorithm changes as the regularization term shrinks .
we actually use algorithmic stability to help us show stability of our algorithm in this sense . theorem 1 states that the regularized least squares regression algorithm is stable as the number of data points increases to infinity .
theorem 2 states that the regularized least squares regression algorithm is stable for a fixed n point data set .
+ * main algorithm * ( regularized least squares regression ) : + for a data set @xmath62 , where @xmath63 , @xmath64 , @xmath65 @xmath66 denote by @xmath45 the function @xmath36 .
denote by @xmath67 the data set @xmath13 .
if @xmath68 and if @xmath69 is chosen to depend on @xmath5 such that @xmath70 and @xmath71 as @xmath6 , then we have convergence of the _ main algorithm _ : @xmath72 here , ` @xmath73 ' denotes convergence in probability .
assume we are given @xmath5 fixed positions @xmath74 .
suppose that for each @xmath75 , the labels are given by @xmath76 , where the @xmath77 s are independent random variables with @xmath78 almost surely .
denote by @xmath79 the data set @xmath13 .
define @xmath4 by : + * ( i ) * @xmath80 for @xmath81 + * ( ii ) * @xmath82 is minimal , among all functions which satisfy * ( i)*. + if @xmath83 is chosen to depend on @xmath84 such that @xmath85 as @xmath86 and @xmath87 , then we have convergence of the _ main algorithm _
almost surely : @xmath88 + section 2 contains a short review of rkhs .
section 3 and 4 contain the proofs of theorems 1 and 2 .
@xmath25 is a real reproducing kernel hilbert space ( rkhs ) if @xmath25 has the following properties : * @xmath25 _ is a hilbert space .
_ @xmath25 is a complete , inner product , real vector space of functions @xmath89 .
we denote @xmath25 s inner product by @xmath90 , and @xmath25 s norm by @xmath26 . *
_ reproducing property .
_ there exists a bilinear form @xmath91 such that @xmath92 , we have @xmath93 and @xmath94 for any @xmath95 .
this @xmath96 is called the ` reproducing kernel ' of the rkhs.@xcite@xcite@xcite .
we sometimes denote @xmath97 by @xmath98 . * _ spanning property .
_ @xmath99 since @xmath25 is a real hilbert space , @xmath100 for all @xmath101 .
it follows that @xmath102 , i.e. @xmath96 is symmetric in its two arguments . an equivalent definition of an rkhs is a hilbert space of functions @xmath21 such that all evaluation functionals @xmath103 , are continuous . given @xmath104 , the associated @xmath105 gram matrix @xmath106 has entries @xmath107 where @xmath96 is the reproducing kernel for the rkhs @xmath25 .
the gram matrix is always a positive semi - definite matrix .
the representer theorem transforms the minimization of our functional @xmath108 into an optimization problem over only @xmath5 numbers .
this advantage is the main reason why scientists take @xmath24 to be a ball in a rkhs @xmath25 .
we present a corollary of this theorem below . + * corollary of the representer theorem * ( kimeldorf , wahba)@xcite ) _ the function @xmath109 can be represented in the form @xmath110 .
this is true for any arbitrary loss function v. _ + ( this corollary is a specific case of the full representer theorem @xcite . ) + having described the basic facts about rkhs , we now continue with the proofs of theorems 1 and 2 .
for the _ main algorithm _ above , we are increasing the size of the data set @xmath67 as @xmath5 increases .
we need to show convergence @xmath111 , where @xmath112 and @xmath6 .
that is , we need to show @xmath113 we can break up the distance @xmath114 into two contributions .
the first contribution is called ` variance ' , and it is due to the finite number of randomly chosen noisy data points .
the variance vanishes with arbitrarily high probability as the number of data points increases , even if the noise does not vanish .
the second contribution is the ` bias ' due to the restriction we place on our hypothesis space , i.e. the fact that @xmath22 is chosen from with a ball of a rkhs .
this term vanishes as the ball gets larger , i.e. when @xmath115 gets smaller .
@xmath116 where @xmath117 where @xmath118 , and @xmath119 , where @xmath120 ^ 2|\mthx)d\mu(\mthx ) + \lambda_n \|f\|_h^2.$ ] lemma 1.1 below describes a method for proving that the minimizers of two convex functions are close in @xmath25 .
+ * lemma 1.1 . * since @xmath121 and @xmath122 are minimizers of @xmath123 and @xmath124 respectively , and using the closeness condition * ( a ) * : @xmath125 + @xmath126 + so , @xmath127 now , @xmath128 . back to the proof of theorem 1 .
we proceed one term at a time .
+ * variance term * we will choose more general versions of @xmath129 and @xmath130 temporarily .
@xmath131 that is , we assume that the loss function @xmath132 is lipschitz continuous , or ` sigma - admissible ' @xcite .
the least squares loss has @xmath133 , since @xmath134 . we need to verify the conditions * ( a ) * and * ( b ) * in order to use lemma 1.1 . to verify the closeness property * ( a ) * for our functionals @xmath135 and @xmath136 : @xmath137
there are many available upper bounds for the right side of equation ( [ rempractual ] ) , including vapnik s vc bound , which relies on the vc dimension of the class of allowed decision functions @xmath24
( @xcite , @xcite ) .
the particular bound we utilize for this paper was constructed by bousquet and elisseeff @xcite , and it is based on ` algorithmic stability ' . in general , bounds of this quantity
are probabilistic , and are based on some capacity measure of the algorithm or space of functions @xmath24 .
this particular bound relies on the sigma - admissibility of the loss function v , and mcdiarmid s concentration of measure inequality .
+ @xmath138 is a training sample @xmath139 .
that is , we replace the @xmath141 training point in @xmath138 by a new data point in order to obtain @xmath142 .
+ the algorithm @xmath143 is _ uniformly @xmath144-stable with respect to loss function _ @xmath145 if : @xmath146 + basically , this algorithmic stability measures how much the algorithm s output could possibly change , as measured by the loss function @xmath147 , when we replace one data point . +
* algorithmic stability theorem * ( bousquet and elisseeff , @xcite ) _ if we are given a uniformly @xmath144-stable algorithm with respect to loss function @xmath147 , which outputs functions bounded by the constant @xmath148 ( i.e. @xmath149 ) , then for any @xmath150 , _ @xmath151 * algorithmic stability of tikhonov learning algorithms * ( bousquet and elisseeff , @xcite ) _ the _ main algorithm _ is uniformly @xmath144-stable , with @xmath152 . here , @xmath153 is an upper bound on the sigma - admissibility constant @xmath154 , and @xmath155 is an upper bound on the diagonal elements of the gram matrix , that is , @xmath156 . _ + returning to the proof of theorem 1 , we now know the right side of ( [ rempractual ] ) is bounded by @xmath157 ( for large values of @xmath5 ) with probability at least @xmath158 , where : @xmath159 we now have the closeness condition * ( a ) * of lemma 1.1 satisfied with probability at least @xmath158 , i.e. @xmath160 , with probability at least @xmath158 , where @xmath161 is given in ( [ thm1p ] ) .
now we verify condition * ( b)*. we need to show that @xmath162 implies that @xmath163 for every @xmath22 .
let s define a function @xmath164 so that @xmath165 in what follows .
@xmath166 d % \mu(\mthx ) % \\ & & + \int h^2(\mthx ) d \mu(\mthx ) + \lambda_n % \|\hat{f}_{\lambda_n}\|_\h^2 + % 2\lambda_n ( \hat{f}_{\lambda_n},h)_\h + \lambda_n\|h\|_\h^2\\ & = & \hat{l}_{\lambda_n}(\hat{f}_{\lambda_n } ) + \big [ 2 \int\e [ ( \hat{f}_{\lambda_n}(\mthx)-y)h(\mthx)|\mthx]d\mu(\mthx ) + \ ; 2\lambda_n ( \hat{f}_{\lambda_n},h)_\h \big ] \\ & & + \int h^2(\mthx ) d \mu(\mthx)+ \lambda_n\|h\|_\h^2\end{aligned}\ ] ] the terms in the brackets are linear in @xmath164 .
remember that @xmath167 is the minimizer of @xmath168 , and thus the linear terms must be zero .
( if the linear terms are non - zero , we can reverse the sign of @xmath164 and contradict @xmath167 as the minimizer . )
the last two terms are always positive .
@xmath169 now we can see that * ( b ) * holds : @xmath170 in this case , @xmath171 since both the conditions * ( a ) * and * ( b ) * are satisfied , lemma 1.1 produces @xmath172 where @xmath161 is given in ( [ thm1p ] ) .
we are done with the variance term .
+ * bias term * we will prove that the bias term vanishes using the spectral theorem .
define the function @xmath173 in what follows .
now , @xmath174d\mu(\mthx ) + \lambda_n % ( h+f_\e , h+f_\e)_\h \\ & = & \int ( k_\mthx , h)_\h^2 d\mu(\mthx ) + \int \e[(f_\e(\mthx ) -
y)^2|\mthx]d\mu(\mthx)+ \lambda_n \|h + f_\e\|^2_\h\end{aligned}\ ] ] the minimizer of @xmath130 again must have first variational derivative equal to 0 .
using fubini s theorem , we find : @xmath175\bigg|_{\gamma=0}\\ % & = & \frac{\partial}{\partial \gamma } \bigg [ \int % ( k_\mthx , h)_\h^2d\mu(\mthx ) + 2\gamma \left(\int ( k_\mthx , h)_\h % k_\mthx d \mu(\mthx),g\right)_\h + \gamma^2 % \int(k_\mthx , g)_\h^2d\mu(\mthx ) \\&&+ \lambda_n % \left[(h+f_\e , h+f_\e)_h + 2\gamma ( h+f_\e , g)_\h + % \gamma^2 ( g , g)_\h\right]\bigg]\bigg|_{\gamma=0}\\ & = & 2\left(\int ( k_\mthx , h)k_\mthx d \mu(\mthx ) + \lambda_n ( h+f_\e),g\right)_\h\end{aligned}\ ] ] if @xmath176 , the above expression must be zero for all @xmath177 , thus : @xmath178 let s define a new operator @xmath179 .
+ @xmath180 + one can check that @xmath179 is self - adjoint since @xmath181 . for an operator @xmath182 from one hilbert space @xmath183 , to another , @xmath184 , the operator norm of @xmath182 is defined by @xmath185 .
our operator @xmath179 is bounded , since by cauchy - schwarz ,
@xmath186 we are going to use the spectral theorem next , but first let us review a few facts from functional analysis about this theorem ( @xcite ) .
the spectral theorem allows one to define functions of a bounded self - adjoint operator on a hilbert space @xmath187 .
if the function is a polynomial , e.g. @xmath188 , then it is clear how to define the corresponding operator @xmath189 .
the spectral theorem extends the correspondence @xmath190 to all continuous functions ( in fact , to all bounded borel functions ) .
moreover , one has @xmath191 . because @xmath192 is a real function , the operator @xmath193 provided by the spectral theorem is also self - adjoint .
in addition , for each @xmath194 , we have a measure @xmath195 on @xmath196 such that @xmath197 .
the measure @xmath195 is concentrated on that part of the spectrum @xmath196 along which @xmath22 has a nonzero component . in particular ,
if @xmath198 , then @xmath22 is an eigenvector of @xmath199 with eigenvalue 0 , and @xmath195 is a @xmath200-measure concentrated on @xmath201 .
if on the contrary , @xmath202 , then @xmath203 . now , using the definition of the spectral measure @xmath204 for the operator @xmath179 and the function @xmath43 , we have from ( [ thm1 t ] ) : @xmath205 since @xmath206 is positive semidefinite , @xmath179 is a positive operator and thus has non - negative spectrum only .
one can see that @xmath207 is empty , i.e. take any function @xmath208 such that @xmath209 .
then , @xmath210 ; thus , @xmath208 must be zero almost everywhere .
it follows that @xmath201 is a set of measure zero for @xmath204 . as @xmath211 , @xmath112 , and
the function @xmath212 converges to 0 pointwise on @xmath213 , and thus almost everywhere with respect to @xmath204 ; since this function is bounded by 1 , we can again use the dominated convergence theorem to say that the integral vanishes as @xmath6 .
one can not give a more explicit bound for this term without more information about the relationship between @xmath214 and @xmath25 . in any case , we have convergence of the bias term to 0 .
now , we can complete the proof of theorem 1 .
for any @xmath215 , we must be able to show that @xmath216 .
so , let us choose an arbitrary fixed value @xmath217 .
the bias term vanishes as @xmath6 and @xmath70 , so there must exist an @xmath218 so that for @xmath219 , @xmath115 is sufficiently small , so that the bias term is bounded by @xmath220 .
thus , we consider the bias term bounded by @xmath220 in the limit as @xmath6 ; since this term does not depend on the data , the bound clearly holds with probability 1 .
now , we must choose @xmath221 so that the variance term is bounded by @xmath220 . using the bound in ( [ thm1variancebd ] ) , we choose @xmath222 .
the corresponding probability @xmath161 is then given by ( @xmath223 ) , @xmath224 we need @xmath161 to vanish as @xmath6 ; this is satisfied if @xmath225 as @xmath211 .
also , there must exist an @xmath218 such that for @xmath219 , we have @xmath226 ; we need this in order to use the algorithmic stability theorem .
thus , theorem 1 is proved .
this section contains the proof of theorem 2 .
first , some notation . the positions @xmath74 will be considered fixed throughout this section .
+ @xmath227 + the ` evaluation operator ' @xmath228 evaluates a function @xmath22 at each position @xmath61 in the data set .
note that @xmath228 ` loses information ' about a function @xmath22 by evaluating it at only @xmath5 points .
that is , @xmath229 is a nontrivial subspace of @xmath25 .
the adjoint @xmath230 of the operator @xmath228 is given by @xmath231 .
one can show that @xmath228 is a bounded operator , with @xmath232 the operator @xmath233 is automatically positive and self - adjoint .
we will later use the spectral theorem on the bounded self - adjoint operator @xmath233 .
we start the proof of theorem 2 with the following lemma . + * lemma 2.1 * : _ the following characterizations of @xmath4 are equivalent _ :
@xmath4 satisfies : + * ( i ) * @xmath80 for @xmath81 , and + * ( ii ) * @xmath234 that satisfy @xmath235 2 .
@xmath4 satisfies : @xmath236 3 .
@xmath4 satisfies : + * ( i ) * @xmath80 for @xmath81 , and + * ( iii ) * @xmath237 we have @xmath238 .
we will show 1 .
@xmath239 2 .
@xmath239 3 .
@xmath239 1 .
@xmath240 2 .
first , we show that the function described in 1 .
is unique . from the reproducing property ,
we know that @xmath4 has nonzero components along each of the @xmath241 s for which @xmath242 .
since @xmath25 is a hilbert space , we can always decompose @xmath4 into a component @xmath243 within the span of the @xmath241 s and a component @xmath244 orthogonal to each @xmath241 ( where @xmath245 ) .
now , @xmath246 .
thus , if @xmath247 , then @xmath4 no longer has minimal norm and contradicts property * ( ii)*. the component of @xmath4 along each of the @xmath241 s is determined by the value of @xmath2 .
so , functions @xmath22 that satisfy both * ( i ) * and * ( ii ) * can be written @xmath248 for the fixed values of @xmath249 , @xmath81 .
in particular , the @xmath249 s must satisfy : @xmath250 thus , the function described in 1 .
is unique .
it is straightforward to see that the function described in 2 .
is exactly the function described in 1 .
evaluating the right side of ( [ lemma2.1 ] ) at @xmath251 , we obtain @xmath252 moreover , the function described in 2 .
lies entirely within the span of the @xmath241 s
. therefore it obeys * ( i ) * and * ( ii ) * and we have 1 .
@xmath240 2 .
@xmath239 3 .
because 2 .
@xmath239 1 . , *
( i ) * is satisfied .
we just need to show * ( iii)*. + for any @xmath253 @xmath254 \3 .
@xmath239 1 . here , *
( i ) * is automatic , so we need to check * ( ii)*. + @xmath255i=1, .. ,n@xmath256 now , @xmath257 thus we have @xmath258 , so lemma 2.1 is proved . back to the proof of theorem 2 .
the functional @xmath259 in the _ main algorithm _ expressed in terms of @xmath228 becomes : @xmath260 the minimizer of @xmath259 must satisfy @xmath261 . in other words ,
the first variational derivative of @xmath259 is 0 at its minimizer .
recalling that @xmath262 , this minimization problem becomes : @xmath263 \big|_{\gamma=0}\\ % & = & \frac{\partial}{\partial \gamma } \big [ \frac{1}{n}(p f - p % \overline{f } - \frac{1}{t}\mathbf{b},p f - p \overline{f } - % \frac{1}{t}\mathbf{b})_{\ell_2 } + 2\gamma\frac{1}{n}(p h , p % f - p\overline{f}-\frac{1}{t}\mathbf{b})_{\ell_2 } + % \\ & & \gamma^2\frac{1}{n}(p h , p h)_{\ell_2 } + \lambda_t ( f , f)_\h % + 2\gamma\lambda_t(f , h)_\h + \gamma^2\lambda_t ( h , h)_\h \big ] \big|_{\gamma=0}\\ & = & 2\frac{1}{n}(p h , p f - p \overline{f } -\frac{1}{t}\mathbf{b})_{\ell_2 } + 2\lambda_t(f , h)_\h \\&= & 2\big(h,\frac{1}{n}p^*p f - \frac{1}{n}p^ * p \overline{f } - \frac{1}{n}p^*\big(\frac{1}{t}\mathbf{b}\big ) + \lambda_t f\big)_\h .\end{aligned}\ ] ]
this must be true for any function @xmath164 , so @xmath264 , implying @xmath265 it follows that @xmath266 in order to show stability , we bound the two terms on the right of ( [ thm2terms ] ) , and construct these bounds so they vanish as @xmath267 .
that is , we need to bound the norms above . to accomplish this
, we will use the spectral theorem on the bounded self - adjoint operator @xmath268 . to bound the first term in equation ( [ thm2terms ] ) , recall that the operator obtained from the function @xmath269 of the self - adjoint operator
@xmath233 is self - adjoint .
also , since @xmath233 is a positive operator , the spectrum @xmath270 of the operator @xmath233 is concentrated on @xmath271 . using the spectral measure @xmath272 on the spectrum @xmath270
, we find : @xmath273 ^ 2 \overline{f } , \overline{f})_\h .
% & = & ( ( \frac{1}{n}p^ * p+\lambda_t)^{-2 } \lambda_t^2\overline{f},\overline{f})_\h \\ \\&= & \int_{spec(p^ * p ) } \left(\frac{\lambda_t}{\frac{1}{n}z + \lambda_t}\right)^2 d \nu_{\overline{f};p^*p}(z)\end{aligned}\ ] ] because @xmath274 , we have @xmath275 for all @xmath276 . by lemma 2.1 , we know that @xmath277 , and since @xmath278 , we know that @xmath279 . since @xmath280 for all @xmath281 , it then follows from the dominated convergence theorem that @xmath282 .
one can not give a more explicit bound for this first term ; it would require more specific knowledge of the relationship between @xmath214 and @xmath25 . in any case
, we have achieved our goal in showing that the first term of ( [ thm2terms ] ) vanishes as @xmath283 .
we need the second term in equation ( [ thm2terms ] ) to vanish also .
recall that for operator @xmath284 , it is true that @xmath285 , and that a continuous real function of a self adjoint operator such as @xmath233 is self adjoint .
@xmath286 we use the spectral theorem for the bounded self - adjoint operator @xmath287 , namely the fact @xmath288 where @xmath289 here .
the maximum value of @xmath290 occurs at @xmath291 , and it is @xmath292 .
thus , @xmath293 as long as we design @xmath294 so that @xmath295 , then we have the desired convergence of this term to 0 .
we are done with the second term of equation ( [ thm2terms ] ) .
theorem 2 is proven .
we have proved stability for the regularized least squares regression algorithm , for the sense in which inverse problems are examined . we have shown stability for this algorithm in two cases : the case when the number of data points @xmath5 is a constant , and the case where @xmath6 .
it is important that our algorithm is stable in this sense , because we do not want any inherent error in the algorithm s output .
neither a small amount of noise in the data nor a small amount of regularization should drastically influence the algorithm s output .
we hope that the reader will gain more from our result than the knowledge that regularized least squares regression is stable in the inverse operator sense .
we have found the particular methods introduced in the proofs of theorem 1 and theorem 2 useful for various learning problems , especially those which require the convexity of learning functionals or convergence of learning algorithms .
namely , we demonstrate two methods for showing that the minimizers of two learning functionals are close : use of the spectral theorem , and the technique of lemma 1.1 , which can both be generally applied to other learning algorithms .
the author would like to express infinite gratitude to ingrid daubechies . | we discuss stability for a class of learning algorithms with respect to noisy labels .
the algorithms we consider are for regression , and they involve the minimization of regularized risk functionals , such as @xmath0 .
we shall call the algorithm ` stable ' if , when @xmath1 is a noisy version of @xmath2 for some function @xmath3 , the output of the algorithm converges to @xmath4 as the regularization term and noise simultaneously vanish .
we consider two flavors of this problem , one where a data set of @xmath5 points remains fixed , and the other where @xmath6 . for the case
where @xmath7 , we give conditions for convergence to @xmath8 ( the function which is the expectation of @xmath9 for each @xmath10 ) , as @xmath11 . for the fixed @xmath5 case , we describe the limiting non - noisy , non - regularized function @xmath4 , and give conditions for convergence . in the process , we develop a set of tools for dealing with functionals such as @xmath12 , which are applicable to many other problems in learning theory .
march 7 , 2003 + |
# 1 quantum topology change in @xmath1 + + @xmath2_department of physics , syracuse university + syracuse , ny 13244 - 1130 , usa _
+ @xmath3_universidade federal de santa catarina , centro de fsica e matemtica , + dep .
mtm , cep 88.010 - 970 , florianpolis , sc , brazil _
+ @xmath4_universidade de so paulo , instituto de fsica - dfma + caixa postal 66318 , 05315 - 970 , so paulo , sp , brazil _
+ it is very common to make the reasonable assumption that the topology of space - time is fixed .
we assume that space - time is a manifold of the form @xmath5 , and that for each , we have a space - like surface that is always homeomorphic to a given @xmath6 .
however , when ( quantum ) gravity is taken into account , the very geometry of space becomes a degree of freedom , and one can conceive the possibility that @xmath6 changes in the course of time @xcite . such a process is called topology change .
creation of baby universes , production of topological defects ( cosmic strings , domain walls ) , and changes in genus ( production of wormholes and topological geons ) are examples of topology change . each of them have received some attention in the literature .
several authors have investigated topology change within the context of both classical and quantum gravity @xcite .
it is interesting to notice that in the usual canonical approach to gravity , only the metric of the spatial manifold @xmath6 appears as a degree of freedom and receives a quantum treatment .
the topology of @xmath6 in its turn is implicitly treated as a classical entity .
there are , of course , other approaches to quantum gravity such as string theory @xcite and euclidean quantum gravity @xcite where topology may appear as an entity of a quantum nature via a sum over topologies
. it would be desirable to have a formalism where topology can in a certain sense be canonically quantized and if possible separated from degrees of freedom coming from metric and other fields . in spite of the fact that topology change has been inspired by quantum gravity , it has been demonstrated in @xcite that it can happen in ordinary quantum mechanics . in this approach
, metric is not dynamical , but degrees of freedom related to topology are quantized .
the notion of a space with a well defined topology appears only as a classical limit .
( see also @xcite for related ideas ) .
the views we would like to present in this paper are similar , to a certain extent , to the ones in @xcite . in our approach , variables related to topology are separated from other degrees of freedom and then quantized . the topology of space is well captured by soliton - like excitations of @xmath6 called topological geons .
they can be thought of as lumps of nontrivial topology .
for example , in , the topology of an orientable , closed surface @xmath6 is determined by the number of connected components of @xmath6 and by the number of of handles on each connected component .
each handle corresponds to a topological geon , i.e. , a localized lump of nontrivial topology .
it is well known that these solitons have particle like properties such as spin and statistics .
however unlike ordinary particles they can violate the spin - statistics relation @xcite .
it has been suggested @xcite that the standard spin - statistics relation can be recovered if one considers processes where geons are ( possibly pairwise ) created and annihilated , but this necessarily implies a change of the topology of @xmath6 . in other words , one may have to consider topology change in order to have a spin - statistics theorem for geons @xcite .
the euclidean path integral approach can in some sense be carried out in low dimensions @xcite , but it represents a formidable task in the case of a @xmath7 theory .
it would be nice to stay closer to a `` canonical '' quantization , even though topology change and the canonical approach appear to be incompatible .
one may search for alternative descriptions of topological properties using algebraic tools , very much in the spirit of quantum invariants of knot theory .
the polynomial invariants of knots can be obtained by both field theoretic and algebraic methods . in the field theoretical approach ,
it is well known that jones polynomials are obtained by means of functional integrals of chern - simons theory @xcite . in the algebraic approach ,
one obtains invariants by representations of the braid group @xcite , or via hopf algebras @xcite .
we will try in this paper to give an algebraic description of quantum geons , rather than a field theoretical one .
we will present a theory of quantized topological geons where topology change is a quantum transition .
we will only analyze the case of orientable geons in @xmath8 were handles are the only possible `` particles '' . a generalization to include nonorientable geons
will be presented elsewhere .
let us consider a manifold @xmath9 and some generic field theory ( possibly with gauge and higgs fields ) interacting with gravity .
it is reasonable to expect that if we could quantize such a complex theory , its observables would give us information on the geometry and topology of @xmath9 .
the main point is that one does not need to consider the full theory to get some topological information .
it is possible that , in a certain low energy ( large distance ) limit , there would be a certain set of observables encoding the topological data .
we know examples where this is precisely the case . in general , the low energy ( large distance ) limit of a field theory
is not able to probe details of the short distance physics , but it can isolate degrees of freedom related to topology
. we may give as an example the low energy limit of @xmath10 super yang - mills , known as the seiberg - witten theory @xcite .
we also have examples of more drastic reduction where a field theory in the vacuum state becomes purely topological @xcite .
inspired by these facts we will identify the degrees of freedom , or the algebra @xmath11 of `` observables '' , capable of describing @xmath12 topological geons in @xmath8 . actually , we will argue later in this paper that the operators in this algebra are not really observables in the strict sense .
rather , it is what is called @xcite a _ field algebra_. we say that @xmath13 describes a single geon in the same way that the algebra of angular momentum describes a single spinning particle . in this framework
what we mean by quantizing the system is nothing but finding irreducible representations of @xmath13 . as in the case of the algebra of angular momentum ,
different irreducible representations have to be thought of as different _ particles_. for the moment , we will not be concerned with dynamical aspects .
we would like to concentrate on the quantization itself and leave the dynamics to be fixed by the particular model one wants to consider .
an intuitive way of understanding the algebra @xmath13 for a topological geon comes from considering a gauge theory with gauge group @xmath14 in two space dimensions spontaneously broken to a discrete group @xmath15 . for simplicity we will assume that @xmath15 is finite . as an immediate consequence it follows that the gauge connection ( at far distances ) is locally flat . in other words ,
homotopic loops @xmath16 and @xmath17 produce the same parallel transport ( holonomy ) .
the set of independent holonomies are therefore parametrized by elements @xmath18 $ ] in the fundamental group @xmath19 .
it is quite clear that such quantities are enough to detect the presence of a handle .
the phase space we are interested in contains only topological degrees of freedom .
therefore such holonomies can be thought of as playing the role of position variables .
we also have to take into account the diffeomorphisms that are able to change @xmath18 $ ] .
they will be somewhat the analogues of translations .
it is clear that the connected component of the group of diffeomorphisms , the so - called small diffeos , can not change the homotopy class of @xmath16 . to change the homotopy class of a curve @xmath16 one needs to act with the so - called large diffeomorphisms .
therefore the analogues of translations have to be parametrized by the large diffeos modulo the small diffeos .
this is exactly the mapping class group @xmath20 .
also , we must take into account an action of the group @xmath15 , changing the holonomies by a conjugation .
this action , as we will discuss later , corresponds physically to `` encircling flux sources at infinity '' .
these three sets of quantities will comprise our algebra @xmath13 .
contrary to what happens in field theory or even in quantum mechanics , we find that @xmath13 is finite dimensional .
this will be important to avoid technical problems of various kinds .
the algebra @xmath13 contains the analogue of positions and translations and can be thought of as a discrete weyl algebra .
there seems to be no great obstacle to generalize our results also to the case where @xmath15 is a lie group @xcite .
our algebraic description of geons is analogous to what has been developed for @xmath21 non - abelian vortices by the amsterdam group @xcite .
these ideas have been further developed by some of us and coworkers and applied to rings in @xmath7 .
their results will not be discussed here since a complete account will be reported in @xcite .
the algebra encountered by @xcite was a special type of hopf algebra , namely the drinfeld double of a discrete group @xcite . in our case , however , the algebra @xmath13 is not hopf , but it has a drinfeld double as a subalgebra . for a pair of geons we find that the corresponding algebra @xmath22 is closely related to the tensor product @xmath23 of single geon algebras .
this fact allows us to determine the appropriate algebra @xmath11 for an arbitrary number @xmath12 of geons .
the main result of our analysis is that it gives us some information on topology change at the quantum level .
this is true for geons as well as for particles on the plane @xcite .
our algebra @xmath13 has to do with large distance observations .
in other words , we can only probe low energy aspects of the theory .
we will argue in section [ s5 ] that geons , i.e , handles in the plane , can be created and annihilated in a quantum fashion as a consequence of the scale of observations .
we would like to mention that other types of topology change , like creation of baby universes , do not fit naturally in our framework and will not be considered here .
one advantage of the algebraic approach is that we can do this analysis without going into the details of the `` complete '' underlying field theory .
we can determine the spectrum @xmath24 of the geons , i.e. , the set of possible irreducible representations of @xmath25 , but a particular field theory may restrict the available possibilities in @xmath24 .
the determination of these possibilities requires the study of particular examples of the underlying field theories .
that may be a very difficult task . in this paper
our intention is to use the simplified algebraic `` field '' theory and see what it can teach us .
it is remarkable that such a simple framework can reveal important features of quantum geons such as a constraint involving spin and statistics as well as rules for quantum topology change .
the former connection is investigated in another paper @xcite .
an approach similar to ours is explored in reference @xcite .
its author views the geon as a vortex - antivortex pair , in which case the algebra describing it _ is _ a quantum double .
this description does not consider the internal diffeomorphisms of the geon , as it aims to describe vortices on a two - dimensional surface with handles .
accordingly , in @xcite , the setting is a two - dimensional surface @xmath26 of _ genus _
@xmath27 _ and @xmath12 punctures _ , whereas in this work we consider a @xmath28-surface @xmath29 of _ genus _
@xmath27 _ without punctures_. our approach is also different inasmuch as we are interested in considering `` large diffeormorphisms '' , i.e. , elements of the mapping class group of @xmath29 .
more specifically , in @xcite , the topology of this surface is a passive background where a theory of pointlike vortices is defined , and its author only deals with diffeomorphisms moving particles ( punctures ) around or through handles . to us , the geons ( handles ) themselves , _ including their internal structure _ , are the entities of interest .
the diffeomorphisms moving these handles are the `` large diffeomorphisms '' we mentioned above . as an illustration of the above mentioned differences , in a typical process considered in @xcite
, one can make a test vortex go through or around a handle , whereas in our case one can conceive of `` test geons '' going through other handles .
our procedure allows a natural generalization towards quantum gravity , which is the issue of another paper @xcite .
we recall the notion of topological geons in section [ s2 ] .
a special emphasis is given to orientable geons in @xmath8 .
the field algebra is described in section [ s3 ] .
section [ gasp ] gives the effective description of a geon as seen from a large distance .
the relevant subalgebra @xmath30 is the same as for a point particle .
the representations of @xmath31 will play an important role when we discuss topology .
quantization of the system is given in section [ s4 ] . in this section
we are able to classify the irreducible representations for a class of algebras @xmath32 that includes our algebra of interest as a particular example .
it is worthwhile to point out that the field algebras for vortices in @xmath8 and for rings in @xmath7 considered in @xcite are also examples of @xmath33 .
section [ s5 ] describes how topology can change in this quantum theory , as a consequence of the scale of observation .
we end with some concluding remarks and prospects of future work . the term _ geon _ was used for the first time by j.a .
wheeler @xcite to designate a lump of electromagnetic energy held together by its own gravitational field , forming a spatial region of non - zero curvature , typically very small . in the context of this paper , however , this term will have a wholly different meaning , namely it will signify a _ topological _ geon , a soliton - like excitation in topology first discussed by friedman and sorkin @xcite , and whose properties were further elaborated by many authors @xcite . in this section , we review the definition and basic properties of topological geons , and refer the reader to the literature for further details .
we start with some basic preliminary definitions .
let @xmath34,@xmath35 be connected @xmath12-dimensional topological manifolds , possibly with boundaries .
we define their connected sum , @xmath36 , as follows : take @xmath12-balls @xmath37 in the interiors of @xmath38 ( @xmath39 ) and remove their interiors .
we thereby add spheres @xmath40 to the boundaries of @xmath34 and @xmath35 .
now identify the points of these spheres via a homeomorphism .
the resulting manifold is @xmath36 .
if @xmath34 and @xmath35 are oriented , we further require that this homeomorphism be orientation reversing so that @xmath36 is also oriented .
it follows trivially from the definition that @xmath41 is homeomorphic to @xmath9 itself .
the connected sum of two tori is shown on fig .
2.1 . in this paper we shall be interested in a decomposition of spacetime by spacelike hypersurfaces ( spatial manifolds ) . in dealing with gravity ,
one is usually interested in spacetime metrics which induce an asymptotically flat ( or asymptotically conical , in the @xmath1
case)riemannian metric on each hypersurface . for a certain `` frozen time '' @xmath43 , the hypersurface @xmath44 should therefore be topologically a manifold with one _ asymptotic region _ , i.e. , there exists a compact region @xmath45 whose complement in @xmath44 is homeomorphic to @xmath46 , where @xmath47 is the dimension of spacetime and @xmath48 is the standard @xmath12-ball in @xmath49 . in fig .
2.2 one can see @xmath1 oriented geons ( which are nothing but handles on a plane , see below ) on a @xmath28-dimensional spatial slice .
this motivates the following definition : an @xmath12-manifold is said to have one asymptotic region iff it is homeomorphic to @xmath50 , where @xmath9 is a closed ( i.e. , compact and boundaryless ) , connected @xmath12-manifold .
typical cases of interest are 2 and 3 manifolds with one asymptotic region , which are to be thought as spatial slices of @xmath0- and @xmath51-dimensional spacetimes respectively .
a plane with a finite number of geons ( handles ) is an example of a manifold with one asymptotic region .
note that all topological complexity can be localized within a circumference @xmath52 , and the geons can be isolated from each other .
outside @xmath52 , one has simply a flat plane . in 2 and 3 dimensions
it is known @xcite that there exists a class of closed connected manifolds @xmath53 called _ prime _ manifolds .
an @xmath12-manifold @xmath9 is said to be prime iff @xmath54 implies that one of @xmath34 , @xmath35 is an @xmath12-sphere .
one can prove that given any compact @xmath12-manifold ( @xmath55 ) @xmath9 , there exists a unique decomposition m = p_1 # ... # p_n , where @xmath56 .
uniqueness means ( apart from some technicalities - see ref .
@xcite ) that given another decomposition @xmath57 , we have @xmath58 and ( after possible reordering ) @xmath53 is homeomorphic to @xmath59 .
each prime component of @xmath9 is called a topological geon . in 2 dimensions , ignoring @xmath60 from consideration , the only prime manifolds are @xmath42 , which is orientable , and the `` cross cap '' @xmath61 , which is non - orientable @xcite . in this paper
we will consider only orientable geons , therefore we will have to deal only with @xmath62 . connected sums with @xmath60 are clearly immaterial . in 3 dimensions
there are infinitely many prime manifolds , only partially classified . as examples
we can give the 3-torus @xmath63 and the `` handle '' @xmath64 . from the aforementioned prime decomposition it is clear that any @xmath12-manifold @xmath9 with an asymptotic region can be decomposed as m=^n # p_1 # ... # p_n . now consider @xmath65 .
one can always find an @xmath66 sphere in @xmath67 whose interior contains @xmath53 . by a suitable choice of the metric this region can be thought of to be as small as one pleases , i.e. , the topological complexity can be localized ( for details see ref .
@xcite ) . in 2 spatial dimensions
this means that one is allowed to put the handle inside of a circle and suppose the radius of the circle to be very small .
then one has a very small handle surrounded by a vast flat plane .
it is in this sense that we refer to the geon as `` soliton like '' at the beginning of this section : just as a soliton corresponds to a localized excitation of some field , outside of which one has the vacuum , the geon is a localized excitation of the topology itself , the `` vacuum '' in this case being the flat space ( see fig .
2.3 ) . in general
, since @xmath53 is prime , one may say that it represents an elementary topological excitation .
we therefore say that @xmath65 is a space with one geon .
the manifold @xmath68 is therefore seen as a space with @xmath69 geons .
these prime manifolds attached to @xmath70 can be isolated from one another in the same way as we localized one single geon @xcite , and for many purposes one can think of geons as particles .
again , in 2 spatial dimensions one can have many isolated small handles .
the importance of geons to us lies in the fact that , as long as we preserve connectivity and consider a space manifold with one asymptotic region , _ topology change amounts to creation and annihilation of geons_. henceforth we restrain our attention to the case when the space is 2 dimensional , connected and oriented , with one asymptotic region .
we assume , furthermore , that connectivity and orientability are preserved during topology change . although somewhat restrictive , this case is still of much interest .
our assumptions imply , on the other hand , that the geons of interest will be those associated to copies of @xmath42 , i.e. , topology changing processes will mean creation and annihilation of handles on a plane . as we will see below ,
creation and annihilation will have for us a meaning different from the usual geometrical one .
instead they will be related to what a `` distant '' observer will be able to measure from a quantum theoretical standpoint .
our aim in this section is to define some `` observables '' which describe the topological character of a geon .
however , the term `` algebra of observables '' to designate the algebra describing geons would actually be a misnomer , for as we will see shortly , this algebra includes operators which can not be observables . to describe geons
, we will use the low - energy limit of a field theory in their presence . in this limit ,
the theory becomes topological , and therefore provides us with quantities capable of probing the topological features of the background , and hence the geons .
the kind of algebra which we will encounter is composed by a part related to the fields , via their holonomies around non - contractible paths , and to physical operations ( some of them not observable locally ) which may change these holonomies . this algebra is what is known in the literature as a _ field algebra _ ( for a detailed definition , see for instance @xcite ) .
we will follow an approach inspired by the work of the amsterdam group , which is reported in ref . @xcite . in this work
, the group investigates the properties of topological solutions of a @xmath8 gauge field theory in minkowski spacetime where the gauge symmetry of a lie group @xmath14 is spontaneously broken to a finite group @xmath15 by a non - vanishing expectation value of a higgs field @xmath71 .
we shall briefly review their discussion , referring the reader to @xcite for details .
the lagrangian is given by = f^a _ f^_a + tr[(d _ ) ^ * ( d^ ) ] - v ( ) , [ higgs ] where @xmath72 , and @xmath73 is a lie algebra index . for simplicity , we assume that @xmath14 is connected and simply connected .
the fields @xmath74 are the components of the field strength of the yang - mills potential @xmath75 and @xmath76 denotes the covariant derivative determined by this potential .
the higgs field @xmath71 is in the adjoint representation and can be expanded in terms of generators @xmath77 of the lie algebra of @xmath14 , and @xmath78 is a @xmath14-invariant potential . in this paper
we shall be concerned with the low energy , or equivalently , the long range behavior of this theory , in the temporal gauge @xmath79 .
this is obtained by minimizing the three terms in the energy density separately . minimizing the term corresponding to the energy density of the yang - mills field
, we obtain the condition @xmath80 , from which we conclude that we are dealing only with flat connections .
the minimum of the potential restricts the values of the higgs field to the vacuum manifold , which is invariant by @xmath15 .
finally , the condition @xmath81 , required for minimizing the energy density from the second term , tells us that the holonomies ( ) = p\ { _ a^a_i t_a ds^i } ; i \{1,2 } , [ flat ] take values in the finite group @xmath15 .
this gauge theory may have topologically non - trivial , static solutions such as vortices .
it is very well known that the core radii of these vortices are inversely proportional to the mass of the higgs boson , and therefore they may be viewed as point - like in the low - energy regime of the theory .
hence , according to a standard argument , to describe the @xmath69-vortex solutions we may consider solutions for the vortex equations [ static ] f^a_ij & = & 0 ; + d_i&= & 0 ; + v ( ) & = & 0 , on a spacetime of the form @xmath83 , where @xmath84 is the plane with @xmath69 punctures , playing the role of the vortices .
now , take a solution @xmath85 for the vortex equations ( [ static ] ) . by fixing a point @xmath86
, the holonomy of @xmath87 around any closed path @xmath88 based at @xmath82 depends only on its homotopy class , since @xmath87 is flat .
it takes values into a subgroup @xmath15 of @xmath14 , which preserves the vacuum manifold , in view of the equations for @xmath71 @xcite .
therefore , any solution of the vortex equations determines a homomorphism @xmath89 , [ homo ] : _ 1 ( ) h , between the fundamental group @xmath90 and the group @xmath15 .
conversely , given such a homomorphism @xmath89 we can define a solution for eqs.([static ] ) in the following way .
take the universal covering space @xmath91 of @xmath84 .
it is the total space of a principal bundle over @xmath84 with structure group @xmath92 . via the homomorphism @xmath89
we can construct an associated principal @xmath15-bundle over @xmath84 , which is a subbundle of the original @xmath14-bundle . since @xmath15 is finite
, this bundle has a unique flat connection @xmath93 , which can be viewed as a reducible connection on the @xmath14-bundle .
we now find a @xmath71 . by fixing some @xmath94 in the vacuum manifold
, we have that , since @xmath71 must be covariantly constant , we define @xmath95 and its value can be obtained for each @xmath96 by parallel transporting @xmath94 along some path from @xmath82 to @xmath97 in @xmath84 : [ transport ] ( x)= p\ { _ p^x a^a_i t_a ds^i } _ 0 .
the pair ( @xmath93 , @xmath71 ) thus constructed is obviously a solution of the vortex equations .
therefore the space of solutions for the vortex eqs .
( [ static ] ) is essentially parametrized by homomorphisms @xmath98 .
each such homomorphism is then a vortex configuration .
let us take the example of a single point - like vortex on the plane . the non - contractible loop @xmath16 that encircles the singularity has holonomy @xmath99 equal to the flux carried by the vortex .
in trying to capture only topological information , one is not concerned with the position of the vortex , but only with its flux , characterized by some group element @xmath100 . in other words , the `` configuration space ''
@xmath101 for the vortex is labeled by elements of @xmath15 .
hence , the algebra @xmath102 of complex - valued functions on @xmath15 with pointwise multiplication plays the role of `` position observables '' .
let us denote by @xmath103 , the characteristic function supported at @xmath100 .
then [ dp ] p_h p_h=_hhp_h . in terms of homomorphisms
, we have not yet exploited all the degrees of freedom the system has .
indeed , we have that @xmath15 can act by conjugating all holonomies based at @xmath82 by an element of @xmath15 .
hence , a flux @xmath104 transforms under @xmath15 as h h^-1 , [ gauge ] in other words , we have an action of @xmath15 by conjugation of the fluxes .
we shall simply refer to this action as the @xmath15-transformations .
the group elements @xmath105 can be regarded as operators , also denoted by @xmath106 , acting on the functions @xmath107 via ( [ gauge ] ) . in other words , [ dg ] hp_h^-1=p_h h^-1 .
the multiplication of two @xmath15-transformations is the same as the group multiplication .
therefore the algebra of such operators turns out to be the group algebra @xmath108 . as for the physical interpretation of the @xmath15-tranformations
we note that the mathematical action depicted in ( [ gauge ] ) is entirely equivalent , from a physical standpoint , to what occurs when one makes a vortex of flux @xmath104 encircle a source of flux @xmath106 at infinity .
since such operation is non - local , one must conclude that the @xmath15-transformations can not be considered local in the theory , i.e. , can not be implemented by local operators .
the field algebra is then the semi - direct product @xmath109 , the so - called drinfeld double .
it has the structure of a quasi - triangular hopf algebra .
the hopf structure @xcite means in particular the existence of a co - product , i.e , a map @xmath110 which is a homomorphism of algebras ( and with further properties to be discussed in section 4 ) . in @xcite the fluxes are seen as particles in @xmath1 and are then first quantized : the ( internal ) hilbert space @xmath111 is constructed , and the elements of the algebra @xmath112 act as operators on this hilbert space .
@xmath111 decomposes into irreducible representations of @xmath112 , corresponding to the different particle sectors of the quantum theory .
the existence of a co - product allows one to understand fusing processes between particles .
the quasi - triangularity implies the existence of the @xmath113-matrix , @xmath114 , responsible for all braiding processes between particles . for further details
see @xcite .
how is the topology of @xmath84 taken into account in this approach ?
first of all , we have seen that the physically distinct vortex configurations are in one - to - one correspondence to the space of conjugacy classes of homomorphisms of @xmath115 into @xmath15 .
moreover , it is well known that for a finite group @xmath15 the latter space is in one - to - one correspondence with equivalence classes of principal @xmath15-bundles over @xmath84 @xcite .
therefore the only degree of freedom in this theory is the topology of these bundles @xcite .
second , a configuration for which the holonomy is trivial around some puncture is indistinguishable , from the standpoint of the low - energy theory , to another in which that particular puncture is absent .
therefore the low - energy theory somehow actually allows for `` topology fluctuations '' of @xmath84 as long as we stay within its limits , and as far as `` creation and annihilation '' of punctures is concerned . in order to determine the field algebra for a topological geon , we will try to follow a method similar to the one used for vortices in the plane , respecting carefully the differences between the two systems .
we will first try to find the analogues of the `` position observables '' for a geon .
now , @xmath84 is the plane with one or more handles , and for simplicity we shall assume throughout the that there are no vortices , i.e. , we work in the zero vortex number sector of the low - energy limit of the theory given by the lagrangian in ( [ higgs ] ) .
this is in constrast with @xcite , where vortices are the central interest .
there , the vortices determine the state of a handle , whereas in the present work all non - trivial configurations will be related solely to holonomies around and through the handles .
in other words , the geons are our main concern , and the background field theory merely defines their states .
the figure shows the loops @xmath120 ( @xmath121 ) .
the homotopy classes @xmath122 $ ] and @xmath123 $ ] generate the fundamental group .
the class @xmath124 $ ] is not independent of @xmath122 $ ] and @xmath123 $ ] .
let us start by taking @xmath6 to be the plane with a handle . on all figures ,
a geon will be thought of as a square hole on the plane , with the opposite sides identified .
one can show that @xmath19 has two generators @xmath125 $ ] and @xmath126 $ ] , shown by fig .
it can be shown that @xmath127=[\gamma_1][\gamma _ 2][\gamma_1]^{-1}[\gamma _ 2]^{-1}.\ ] ] actually , @xmath19 is freely generated by @xmath125 $ ] and @xmath126 $ ] .
let @xmath128,[\gamma _
2])\in \pi _
1(\sigma ) $ ] , be a word in @xmath125,[\gamma _ 2]$ ] and their inverses .
then @xmath129 maps @xmath27 to @xmath130 where @xmath131 and @xmath132
. therefore the map @xmath133 is completely characterized by the fluxes @xmath134 and @xmath135 . since there is no relation between @xmath73 and @xmath136 , the set @xmath101 of all maps is labeled by @xmath137 .
* definition * : let @xmath15 be a finite group and @xmath84 the plane with one geon , i.e. , a two dimensional manifold given by @xmath138 let @xmath139 and @xmath140 denote representative loops whose classes generate @xmath90 .
we define a classical configuration @xmath141 of a geon as the homomorphism defined by _ ( a , b ) ( _ 1)=a , _ ( a , b)(_2)=b . it is important to bear in mind that @xmath142 and therefore that it is a finite discrete set . for simplicity of notation
, a geon configuration will be denoted simply by a pair @xmath143 of fluxes .
note that we are not explicitly identifing those configurations which differ by an @xmath15-transformation .
this is because wave functions need only be `` covariant '' under the symmetries of the problem , and only its modulus squared and other observable quantities , like aharonov - bohm phases , must be invariant . in our approach , this will happen naturally , just as in @xcite . with @xmath144 being the configuration space for a geon ,
the corresponding algebra of `` position observables '' is @xmath145 . instead of working with the abstract algebra , we specify a representation .
let @xmath146 be the ( finite - dimensional ) complex vector space generated by the vectors @xmath147 .
we will call the representation on @xmath146 , to be defined below , the defining representation .
the algebra @xmath148 is generated by projectors on @xmath146 denoted by @xmath149 . they are defined by [ projectors ] q_(a , b ) c , d = _ a , c _ b , d c , d . the operator @xmath149 represents a `` delta function '' supported at @xmath143 ,
i.e. , it gives @xmath150 when evaluated on @xmath143 , and zero everywhere else .
indeed , from ( [ projectors ] ) one finds that [ algebraf ] q_(a , b ) q_(c , d ) = _ a , c _ b , d q_(c , d ) , besides the projectors @xmath149 , which play the role of position operators in ordinary quantum mechanics , we have also some operators capable of changing @xmath143 .
they are somewhat analogous to momentum operators .
for example , like in the case of vortices , @xmath15-transformations act on the configurations .
it turns out that for a geon there are additional operators besides @xmath15-transformations .
they correspond to the action of the group @xmath151 of diffeomorphisms of @xmath6 that keeps infinity invariant .
the group @xmath15 acts on @xmath101 simply by conjugating both fluxes in @xmath143 .
this will induce an operator @xmath152 for each @xmath153 , acting on the defining representation @xmath146 by _
g a , b = gag^-1 , gbg^-1 . [ del ] from ( [ del ] ) one sees that the multiplication of operators @xmath152 is given by [ groupalgebra ] _ g _ h = _ gh . the corresponding algebra generated by @xmath154 is the group algebra @xmath155 .
the relation between @xmath156 and @xmath155 can be derived from ( [ projectors ] ) and ( [ del ] ) .
one sees immediately that [ actionofh ] _
g q_(a , b)^-1 _ g = q_(gag^-1 , gbg^-1 ) . in other words ,
the algebra @xmath155 acts on @xmath156 . besides @xmath15-transformations , fluxes
@xmath143 can change under the action of the group @xmath157 .
it is clear that elements belonging to the subgroup @xmath158 , the component connected to identity , act trivially on @xmath19 to be at infinity . even if we do not , the holonomies will be invariant under the action of @xmath158 . ] and hence on @xmath143 .
therefore what matters is the action of the so - called mapping class group @xmath20 @xcite , defined as [ mcg ] m _ = . for the present case , @xmath6 is the plane with a single geon and the mapping class group is isomorphic to the central extension of the group @xmath159 , denoted by @xmath160 and called the steinberg group .
this is the same as the mapping class group of a torus minus one point @xcite .
we denote generators of @xmath161 by @xmath87 and @xmath162 .
they correspond to ( isotopy classes of ) diffeomorphisms called dehn twists .
a dehn twist is realized as follows .
take a loop in @xmath84 .
then draw an annulus enclosing the loop and introduce radial coordinates @xmath163 $ ] , with @xmath164 and @xmath165 corresponding to the boundaries of the annulus , see fig .
then rotate the points of the annulus in such a way that the angle of rotation @xmath166 is zero for @xmath164 and gradually increases , becoming @xmath167 at @xmath165 .
figure 3.2 shows how to produce dehn twists , and in fig .
3.3 , we show how the dehn twist @xmath162 deforms the loop @xmath117 .
there is also the dehn twist along a loop enclosing the geon , which can be interpreted as the @xmath168-rotation of the geon @xcite .
this dehn twist will be important when we discuss the notion of spin of a topological geon .
the corresponding annulus is denoted by @xmath169 in fig .
however , @xmath169 is not independent of @xmath87 and @xmath162 .
one can show that @xcite c_2=(ab^-1a)^4 .
the group @xmath20 is generated by @xmath87 and @xmath162 , with the relation that @xmath169 commutes with @xmath87 and @xmath162 .
it is useful to think of the elements of @xmath20 as words @xmath170 in @xmath87 , @xmath162 and their inverses .
the action of @xmath87 and @xmath162 on @xmath171\in \pi_{1 } ( \sigma ) $ ] induces an action on @xmath172 , and therefore induces operators @xmath173 and @xmath174 in the defining representation acting on @xmath146 .
let us take as an example the action of @xmath162 on @xmath139 , as given by fig .
one sees that @xmath125 \rt [ \gamma _ 1][\gamma _ 2]$ ] , and therefore @xmath175 . on the other hand
, @xmath162 keeps @xmath126 $ ] invariant .
one can verify that @xmath87 and @xmath162 induce the following operators : [ abc ] a a , b & = & a , ba , + b a , b & = & ab , b .
for an arbitrary word @xmath170 , the corresponding operator is @xmath176 , i.e. , the same word but with @xmath87 and @xmath162 replaced by @xmath173 and @xmath174 .
for example , the dehn twist @xmath169 of fig .
3.2 is written as @xmath177 and the corresponding operator @xmath178 can be immediately computed to be [ opc ] c_2 a , b = ^-1a*c * , * c*^-1b*c * , where @xmath179 .
the group @xmath182 generated by @xmath173 and @xmath174 defined by ( [ abc ] ) is the one relevant for defining the field algebra .
contrary to the mapping class group , @xmath183 is a finite group .
it turns out that an infinite number of words @xmath176 is equal to the identity operator and that @xmath183 can be naturally identified with @xmath184 divided by a certain normal subgroup .
let @xmath185 be a subgroup of @xmath160 defined as @xmath186 it is easy to see that @xmath185 is a normal subgroup .
in fact , given any word @xmath187 , such that @xmath188 we have the relation w^-1(a , b ) h w(a , b ) a , b & = & w^-1(a , b ) h a , b = w^-1(a , b ) a , b = + & = & a , b .we define the _ effective _ mapping class group @xmath183 acting on the defining representation @xmath146 as the quotient @xmath189 the algebra generated by the operators @xmath173 and @xmath174 is the group algebra @xmath194 .
together with @xmath155 and @xmath156 it gives us the total field algebra @xmath13 for a single topological geon . from the definitions ( [ projectors ] ) , ( [ del ] ) and ( [ abc ] ) one sees that @xmath195 @xmath196 @xmath197 aq_(a , b)a^-1=q_(a , ba ) , bq_(a , b)b^-1=q_(ab , b).[relations ] therefore , both algebras @xmath155 and @xmath194 act on @xmath156 .
the action of a generic word @xmath176 on @xmath149 will be denoted by [ waction ] w(a , b)q_(a , b)w^-1(a , b)= q_(w^(a),w^(b ) ) . where @xmath198 is a pair of words in @xmath73 and @xmath136 and their inverses , representing the action of @xmath170 on @xmath143 .
there are two equivalent ways of presenting @xmath13 .
one is by using the defining representation of ( [ projectors ] ) , ( [ del ] ) and ( [ abc ] ) .
another way is to define @xmath13 as the algebra generated by @xmath149 , @xmath199 , @xmath173 and @xmath174 with the relations ( [ relations ] ) . in any case
, we have that [ productina ] a^(1 ) = ( h ) ( hh ) .
we shall now introduce the field algebra for two topological geons following exactly the same ideas as for a single topological geon .
we recall that for a single geon , @xmath13 consists of three sub - algebras , namely the `` position observables '' @xmath148 , the @xmath15-transformations @xmath155 , and the `` translations '' , i.e. , a realization @xmath183 of the mapping class group @xmath20 .
the algebra @xmath22 for two geons will consist of the same three distinct parts , with @xmath200 and @xmath6 replaced by a plane with two handles .
we shall start by examining the fundamental group @xmath201 let @xmath202 be the loops shown by fig .
one can show that @xmath19 is the free group generated by @xmath171 $ ] . a `` configuration '' @xmath129 of two topological geons
is given by a homomorphism @xmath203 .
therefore @xmath129 is completely characterized by the holonomies @xmath204 along the loops @xmath205 .
since there are no relations among @xmath171 $ ] s , the holonomies @xmath206 and @xmath207 are four arbitrary elements of @xmath15 . in other words , the set @xmath208 of configurations @xmath129 can be identified with @xmath209 , where @xmath210 is the configuration space for a single geon .
the corresponding algebra @xmath211 is thus the direct product of the algebra of single geons , i.e .
, @xmath212 it is natural to work with the defining representation on @xmath215 spanned by vectors of the form @xmath216 where the subscripts denote the respective geons .
the `` position observables '' are generated by projectors @xmath217 acting on @xmath215 in the obvious way , i.e. , q_(a_1 , b_1 ) q_(a_2 , b_2 ) a_1 , b_1 a_2 , b_2 = _
a_1 , b_1 a_2 , b_2 .
[ position ] therefore , the `` position '' operators belong to @xmath23 .
the action of @xmath15-transformation @xmath153 on the fluxes @xmath218 is by a global conjugation .
this induces the action a_1 , b_1 a_2 , b_2 ga_1g^-1 , gb_1g^-1 ga_2g^-1 , gb_2g^-1 on @xmath215 .
the corresponding operator is obviously identified with @xmath219 , since _
g_g a_1 , b_1 a_2 , b_2 = ga_1g^-1 , gb_1g^-1 ga_2g^-1 , gb_2g^-1 .
[ 2gauge ] hence , @xmath15-transformation operators also belong to @xmath23 .
we now start to consider the action of the mapping class group @xmath20 . for two or more geons ,
@xmath20 is much more complicated than for a single geon @xcite .
the mapping class group is generated by dehn twists of the type @xmath87 and @xmath162 ( see fig .
3.2 ) for each individual geon together with diffeomorphisms involving pairs of geons . let @xmath220 , @xmath39 be the generators of the `` internal diffeos '' for each individual geon .
the corresponding operators acting on @xmath215 are clearly given by @xmath221 b_1=b , b_2=b [ internal ] where @xmath222 is the identity operator on @xmath146 .
there are two additional classes of transformations besides the internal diffeos .
the first one , called exchange , is the analogue of the elementary braiding of two particles .
the second , called handle slide , has no analogue for particles , since it makes use of the internal structure of the geon .
so far , all operators in the algebra for @xmath22 were of the form @xmath223 .
it turns out that this is not the case for exchanges and handle slides .
they correspond somewhat to interactions and can not be written strictly in terms of operators in @xmath23 . in order to describe interactions between geons ,
we need to define a pair of flip automorphisms of @xmath224 .
they are necessary in the construction of the exchange and handle slide operators .
* definition * : given a two geon state @xmath225 the flip automorphisms @xmath226 and @xmath16 are defined by : a_1 , b_1 a_2 , b_2 & : = & a_2 , b_2 a_1 , b_1 , + a_1 , b_1 a_2 , b_2 & : = & a_1 ,
b_2 a_2 , b_1 .
both are not given geometrically as morphisms of the mapping class group , but unless one introduces these operators , the algebra of two geons can not be related directly to the algebras for a single geon .
we will show that the algebra @xmath22 can be obtained from the tensor product @xmath23 when we add @xmath226 and @xmath16 .
the effect of a geon exchange on the states is of the form [ geonexchange ] r a_1 , b_1 a_2 , b_2 = _ 1 ^ -1 a_2 * c*_1 , * c*_1 ^ -1 b_2 * c*_1 a_1 , b_1 , where @xmath227 .
this operator is equivalent to braiding operators for particles and also satisfy the yang - baxter equation , [ yb ] ( r ) ( ) ( r ) = ( ) ( r ) ( ) .
one can verify that the exchange operator ( [ geonexchange ] ) may be written as the product [ r ] r= r where @xmath228 is the analogue of the universal @xmath113-matrix for a quasi - triangular hopf algebra . in our case
@xmath113 is given by [ r ] r=_a , bq_(a , b)^-1 _ aba^-1b^-1 .
the handle slide is shown in fig .
3.6 . in ( a )
, the geon is viewed as a rectangular box on the plane . in ( b ) , we have identified two edges of the rectangle and the geon is represented as two circles on the plane connected by dotted lines . the handle slide is defined as the operation that performs a double counterclockwise exchange of the 2nd and 3rd circles followed by a clockwise @xmath229-rotation of each one of them . the handle slide is interpreted geometrically as the full monodromy of two handles followed by a rotation of @xmath229 of each handle .
the figure shows two equivalent representations for the handle slide : in ( a ) , the geon is viewed as a rectangular box on the plane . in ( b ) , we have identified two edges of the rectangle and the geon is represented as two circles on the plane .
as expected , this dehn twist acts on the generators @xmath171 $ ] of @xmath19 given in fig .
3.4 , and therefore on the holonomies . under the action of the transformation indicated in fig .
3.6 , the loops @xmath171 $ ] will be mapped into new loops @xmath235 $ ] . as an example
let us show how the handle slide acts on @xmath139 .
the loop @xmath139 is shown in fig .
3.7 ( a ) . after the action of the diffeo
, @xmath139 is mapped to @xmath234 , indicated in fig .
3.7 ( b ) .
we need to express @xmath234 in terms of the generators @xmath125 $ ] .
it is easy to see that @xmath234 can be deformed to @xmath236 , where @xmath237 is the loop enclosing the second geon .
the sequence of deformations is indicated by fig .
3.7 ( b ) , ( c ) and ( d ) .
as @xmath237 measures the total flux @xmath238 , it is easy to see that @xmath234 will measure the flux @xmath239 .
one can repeat the same procedure for the other loops and show that the action on the loops induces an action on @xmath215 given by the following handle slide operator @xmath240 : [ handleslide ] & & s a_1 , b_1 a_2 , b_2 = + & & = a_1 ( a_2 b_2 a_2 ^ -1 ) , ( a _ 2 b_2 a_2 ^ -1)^-1 b_1 ( a_2 b_2 a_2 ^ -1 ) ( a_2 b_2 a_2 ^ -1)^-1 b_1 ( a_2 b_2 a_2 ^ -1 ) a_2 , b_2 .
this is a very complicated action on states , but there is a way to write @xmath240 as a product of elements of @xmath241 with flip automorphisms in the same way as the operator @xmath242 .
the result is [ h ] s & = & ( _ g , hq_(g , h ) _ g ) ( b ) + & & ( c _ ) ( _ g , hq_(h , g ) _
g^-1 ) + & & ( c _ ) ( b^-1 ) ( c_- ) . these constructions can be easily generalized to write down the algebra @xmath11 for @xmath12 geons .
it is clear that @xmath245 the complete algebra @xmath11 can be obtained by adding the operators @xmath246 and @xmath247 of exchange and handle slide between the @xmath248-th and the @xmath249-th geons .
they can be easily constructed by using operators analogous to ( [ r ] ) and ( [ h ] ) , acting on the @xmath248-th and @xmath249-th entries of @xmath250 .
it is clear that the elements @xmath251 , @xmath252 , @xmath246 and @xmath253 of @xmath11 generate , under multiplication , a group @xmath254 that is homomorphic to the mapping class group @xmath20 for @xmath12 geons . besides the relations proper to @xmath20
, however , we will have extra relations so that @xmath254 becomes effectively finite .
we have seen up to now that a geon is a topological object with internal structure . in quantum theory , it can be described by the algebra @xmath13 . however , for a large distance observation , we may disregard the operators that probe its internal structure and describe it by a subalgebra @xmath255 . in this approximation
, a topological geon seems to be no different from a particle on the plane , or a vortex in @xmath8 .
we may guess that the large distance field algebra @xmath255 is an algebra equivalent to @xmath112 , the quantum double introduced in section [ s3 ] .
actually this is not exactly true .
we will see that @xmath255 for a single geon has extra elements besides the ones corresponding to @xmath112 .
long distance observables should not see the internal structure of the geon .
for instance , in performing aharonov - bohm - type experiments in this long - distance scale , one should expect to see only the effects of the total flux , or the holonomy of the large loop @xmath256 .
therefore , the only detectable projector in this scale is the one with support at the total flux @xmath257 of a single geon .
it is naturally defined as [ qc ] q^(1)_c : = _ a , b _
aba^-1b^-1 , c q_(a , b ) .
the index @xmath258 in @xmath259 is to remind us that this large distance projector is an element of @xmath260 , the algebra of a single geon .
the algebra of operators @xmath259 can easily be obtained from the algebra ( [ algebraf ] ) , resulting in [ subalgebra ] q^(1)_c_1 q^(1)_c_2 = _ c_1 ,
c_2 q^(1)_c_1 . hence
, the algebra generated by @xmath261 is isomorphic to @xmath102 .
the @xmath15-transformation operators @xmath262 act on @xmath263 . from ( [ relations ] ) and ( [ qc ] )
, one can verify that [ subgauge ] _ gq^(1)_c_g^-1=q^(1)_gcg^-1 .
therefore @xmath199 has to be regarded as a large distance operation . to make the notation uniform , we define ^(1)_g:=_g .
the operators @xmath261 should commute with local operators in @xmath13 , namely the diffeos @xmath173 and @xmath174 .
this must be true since the action of the mapping class group can not change @xmath124 $ ] .
that is because one can make @xmath264 very large , such that the dehn twists @xmath87 and @xmath162 do not act on @xmath264 .
3.1 and fig . 3.2 .
in fact , one can verify that aq^(1)_ca^-1=bq^(1)_cb^-1=q^(1)_c .
let us call @xmath265 the algebra generated by @xmath261 and @xmath266 .
it is clear that @xmath267 is isomorphic to the drinfeld quantum double @xmath268 . as a consequence
, @xmath269 has the structure of a quasi - triangular hopf algebra @xcite . in this paper
we will be interested mostly in two properties of a quasi - triangular hopf algebra , namely the existence of a co - product and the universal @xmath113 matrix .
a co - product on @xmath267 is a linear map @xmath270 which is co - associative , @xmath271 and a morphism of algebras , @xmath272 for the quantum double , the co - product has the expressions [ coproduct1 ] ( q^(1)_c ) = _ g q^(1)_gq^(1)_g^-1c . and [ coproduct2 ] ( ^(1)_g)=^(1)_g ^(1)_g .
the quasi - triangularity of the quantum double implies the existence of an @xmath113-matrix , which is responsible for the exchange processes .
the @xmath113-matrix for the quantum double can be written as [ rsub ] r^(1 ) = _ gh q^(1)_g ^(1)_g^-1 .
we recall that the full algebra @xmath13 also has an @xmath113-matrix given by ( [ r ] ) .
one should ask whether the @xmath113-matrix ( [ rsub ] ) for the subalgebra @xmath265 is compatible with ( [ r ] ) .
it is a simple matter to show that they are actually identical .
we may think of the @xmath113-matrix for @xmath13 as a trivial extension of the @xmath113-matrix of @xmath267 .
an important question is whether it is also possible to extend the co - product to the entire algebra @xmath13 .
we have reasons to believe that @xmath273 can not be extended .
one reason is that the co - product is related to fusion of particles at the quantum level , which is physically reasonable .
however , it is harder to imagine that two handles put together could be seen as a single handle .
another large distance element in @xmath274 is the operator @xmath275 responsible for the dehn twist on a cycle that encloses the entire geon . in other words
, @xmath275 is the @xmath229-rotation of the geon : c^(1 ) c_2 .
note that @xmath275 commutes with all elements of @xmath269 .
since @xmath276 for some @xmath69 , it generates a group algebra isomorphic to @xmath277 .
consider next the two - geon configuration and its corresponding algebra @xmath280 .
the associated long distance algebra can be visualized as follows .
let the two geons shrink to localized objects and at the same time approach each other . at the end
a point - like object will remain and we should look for the operators that still make sense in the limit .
it is clear that such operators will be a ) the total flux projector @xmath281 of the two geons ; b ) the @xmath15-transformations and c ) the dehn twist around a cycle enclosing both geons .
the projection operator for the total flux of the system is given by q^(2)_c : = _ a , b , a,b _ aba^-1b^-1aba^-1b^-1 , c q_(a , b)q_(a,b ) .
the index @xmath282 indicates that @xmath281 is an element of @xmath22 .
one can write this expression in a more transparent way as follows : [ q2 ] q^(2)_c=_gq^(1)_gq^(1)_g^-1c .
similarly , the @xmath15-transformation is given by [ d2 ] _
g ^(2):=^(1)_g^(1)_g .
if we compare the last two equations with the definition ( [ coproduct1])-([coproduct2 ] ) , we see that [ q2 ] q^(2)_c=(q^(1)_c ) , [ g2 ] ^(2)_g=(^(1)_g ) .
let us denote by @xmath283 the algebra generated by @xmath281 and @xmath284 . from ( [ q2 ] ) and ( [ g2 ] ) it follows that @xmath283 is homomorphic to @xmath267 .
actually , it is a simple matter verify that they are isomorphic . as in the previous case
, the long distance algebra @xmath285 has an extra generator given by the dehn twist @xmath286 on a cycle enclosing both geons , with c^(2)=r^2 .
as one would expect , the algebra @xmath285 is isomorphic to @xmath279 and therefore also describes a particle with a frame .
it is clear now what is the long distance algebra @xmath287 for @xmath12 geons .
it is generated by the dehn twist @xmath288 on a cycle enclosing the @xmath12 geons , together with elements @xmath289 given by the iterative application of the co - product .
for example , for @xmath290 , q_c^(3 ) & = & ( id)(q^(1)_c ) , + _
g^(3 ) & = & ( id)(^(1)_g ) . notice that because of the co - associativity property , we could have written @xmath291 instead of @xmath292 in the last two formulae .
the algebra @xmath13 describes the topological degrees of freedom for a single geon on the plane . to quantize the system we need to find an irreducible representation of @xmath13 on a hilbert space @xmath111 .
however , this hilbert space will branch into irreducible representations of the field algebra : = _ rh_r , where @xmath293 denotes a particular irreducible representation describing a certain geon type .
the algebra is finite dimensional , and therefore there will be a finite number of irreducible representations of @xmath13 .
furthermore , the hilbert spaces @xmath111 are all finite dimensional .
each representation gives us a possible one - geon sector of the theory . in the case of quantum doubles
, the irreducible representations are fully classified .
see for instance ref .
for the case of geons , the algebra is more complicated because of the existence of internal structure . nevertheless , the representations of @xmath13 are quite similar to the ones of the quantum double of a finite group .
this is not totally surprising , since in a certain limit , as discussed in the previous section , we recover the quantum double @xmath294 .
actually , we can define a class of algebras @xmath295 that can have its representations classified and that are generic enough to contain the quantum double and the algebra @xmath13 as particular cases . in the spirit of @xcite , one can then get all representations of @xmath295 . *
definition * : let @xmath296 be a finite set and @xmath14 a finite group acting on @xmath296 . in other words , there is a map @xmath297 for each @xmath298 . as usual , we denote by @xmath299 the algebra of functions on @xmath296 and by @xmath300 the group algebra of @xmath14 .
we define the algebra @xmath295 as the vector space @xmath301 with basis elements denoted by @xmath302 , @xmath303 and @xmath304 , and the multiplication [ prodina ] ( q_x , g ) ( q_y , h ) : = ( q_x q__g ( y ) , g h ) . here , @xmath305 is the characteristic function supported at @xmath306 .
let @xmath307 be an element of @xmath296 .
we denote by @xmath308 the stability subgroup with respect to @xmath307 , i.e. , k_x_0 = \ { gg _ g ( x_0 ) = x_0 } . the stability subgroup @xmath309 divides the group @xmath14 into equivalence classes of left cosets .
let @xmath69 be the number of equivalence classes and let us choose a representative @xmath310 , @xmath311 for each class , with the convention that @xmath312 .
we can write the following partition of @xmath14 into left cosets : g = _ 1 k_x_0 _ 2 k_x_0 _ n k_x_0 .
let us point out that @xmath313 seen as a vector space carries a left representation of @xmath14 , the action of @xmath14 being by left product .
all irreducible representations of @xmath14 can be obtained by reducing this representation . in particular ,
any vector space carrying an irreducuble representation ( irr ) of @xmath309 can be viewed as a subspace of @xmath313 .
note that @xmath299 plays a dual role : it is an algebra , but it is itself a vector space which is acted upon by the group @xmath14 , according to @xmath314 .
this can be extended to an action of @xmath313 in the obvious way . also , it acts upon itself by left ( pointwise ) product . in what follows
we shall denote the elements @xmath305 by @xmath315 whenever we want to view it as a vector belonging to the representation of @xmath313 on @xmath299 just defined . in this `` passive '' role
it is acted upon , instead of acting on some representation of the algebra of functions .
* theorem * let @xmath316 , @xmath317 be a basis of a subspace @xmath318 of @xmath313 carrying an irr @xmath319 of @xmath309 .
then , for ( a fixed ) @xmath320 , elements @xmath310 , @xmath311 and @xmath321 , @xmath317 as stated above , the vectors @xmath322 form a basis for an irr of the algebra @xmath295 , given by @xmath323 where @xmath324 and @xmath325 are uniquely determined by the equation @xmath326 and @xmath327 is the matrix for the representation @xmath319 .
the quantum double @xmath112 and the algebra @xmath13 are particular cases of @xmath295 .
the quantum double is obtained by taking @xmath328 , @xmath329 , with the action @xmath330 . as for the algebra of a single geon , one takes @xmath331 and for the group @xmath14 the product @xmath332 .
the actions of @xmath333 and @xmath334 commute and are given by @xmath335 and @xmath336 where we have used the notation of ( [ waction ] ) .
the irr s for the algebra ( [ productina ] ) can be constructed given an element @xmath337 .
the stability subgroup @xmath338 is defined by k_(a , b ) = \ { ( g , w ) h _ g_w ( a , b):=(gw^(a)g^-1,gw^(b)g^-1)=(a , b ) } .
then , after choosing representatives @xmath339 for the cosets , the partition of @xmath332 can be written as [ k ] h = _ 1 k_(a , b ) _ 2 k_(a ,
b ) _ n k_(a , b ) . let @xmath340 be a basis of an irr of @xmath341 . then , according to the theorem , the vectors [ irep ] _
_ i(a , b ) j _ , with @xmath311 , @xmath317 , form a basis of an irr of the algebra @xmath13 .
let us express the representations of @xmath13 in a more compact notation .
the action of @xmath332 on @xmath342 divides @xmath296 into orbits .
we denote by @xmath343 $ ] the orbit containing the element @xmath344 .
we will collectivelly call @xmath345 the quantum numbers labeling the irr s of @xmath341 .
one can see from ( [ irep ] ) that an irr @xmath346 is characterized by a pair @xmath347,\rho ) $ ] . a basis for an irr @xmath346 of @xmath13
will therefore be written as vectors @xmath348 , @xmath349 ; @xmath350 defined by [ irepvec ] i , j^(a , b)_r:=_i a , b j _ where @xmath351 is a state in the defining representation , @xmath352 are the same as in ( [ k ] ) and @xmath353 are base elements in the irreducible representations @xmath345 of @xmath341 .
of course , the set of vectors thus defined depend on the pair @xmath143 we choose .
we fix an @xmath73 and a @xmath136 , and henceforth omit the superscript .
let @xmath355 be a generic element of @xmath332 . the equation [ gaction ] _
gw_i=_i defines uniquely a new class @xmath324 , together with an element of the stability group @xmath356 .
the action of @xmath357 on @xmath358 is determined by ( [ gaction ] ) and it reads [ aaction ] _ gwi , j_r & = & _ ia , b j_= + & = & _ k^()()_kj i,k_r where @xmath359 is the matrix representation of @xmath341 .
each irr @xmath347,\rho ) $ ] describes a distinct quantum geon .
the corresponding vector spaces @xmath360 generated by states @xmath361 , are all finite dimensional
. therefore we can easily make it into a hilbert space by introducing the scalar product i,ji , j _
r=_ii_jj. since the algebras @xmath13 are not the same for different choices of the discrete group @xmath15 , we can not say in general what is the spectrum of a geon . first , we need to fix a group @xmath15 and then compute the spectrum for the corresponding @xmath13 .
consider now two geons described by representations @xmath362 and @xmath363 .
the associated hilbert space of states is simply [ 2geonshilbert ] h^(12):=h_r_1_r_2 . as explained in section [ s3 ] ,
the field algebra consists of @xmath23 together with @xmath242 and @xmath240 .
the elements of @xmath23 act naturally on ( [ 2geonshilbert ] ) .
it remains to be said what is the action of @xmath242 and @xmath240 on states in @xmath364 .
we may think of @xmath242 and @xmath240 as scattering matrices for a pair of geons .
the @xmath242-matrix represents an `` elastic '' interaction in the sense that two incoming geons of quantum numbers @xmath362 and @xmath363 are scattered into two objects carrying the same quantum numbers @xmath362 and @xmath363 .
the handle slide @xmath240 on the contrary is a nontrivial scattering , each one of the two outgoing geons being a superposition of many geons in the spectrum . in this paper
we have considered @xmath8 manifolds such that any spatial slice consists of a plane with a certain number @xmath12 of handles . in other words , for each fixed time , the configuration consists of @xmath12 geons .
if the number of geons is not fixed , we say that topology can change .
creation of baby universes is also a topology - changing process , but we will not consider it here for reasons that should become clear in what follows .
our system is described by a certain field algebra @xmath11 , and its quantization is given by a representation of @xmath11 .
a change in the number of geons means necessarily a change in the field algebra .
let us see how that can be accomplished .
let us suppose that a geon , represented by a square with opposite sides identified , has a typical size @xmath366 that can vary with time .
intuitively , a geon can disappear if @xmath366 becomes too small . in this case
, a geon will resemble a point - like object .
let us consider the limiting case @xmath367 .
it is clear that the holonomies associated with loops @xmath139 and @xmath140 of fig .
3.1 do not make sense in this limit .
the only flux observable available in this limit is the holonomy of @xmath264 , responsible for measuring the total flux .
the algebra describing the limiting situation is the one - particle approximation @xmath368 introduced in section [ gasp ] .
actually , we do not need to consider the limit @xmath367 , since our description is supposed to be an effective theory that is not valid beyond a certain scale of energy ( distance ) .
we may say that after the geon has become very small , the operators associated with individual fluxes no longer belong to the low energy ( large distance ) description .
the quantum theory is described by states belonging to an irr @xmath346 of @xmath13 . from the inclusion i^(1):a_l^(1)^(1 ) , it follows that @xmath346 is also a ( in general reducible ) representation of @xmath369 .
let @xmath360 be the vector space carrying the representation @xmath346 .
in general , @xmath360 is decomposable as a direct sum [ decomp ] h_r=_n^r_v _ , n^r _ where @xmath370 carries the irr @xmath104 of @xmath255 .
the long distance observer does not see operators mixing different irr s .
therefore , a long distance observer interprets ( [ decomp ] ) as saying that a geon carrying a representation @xmath346 can decay into different _
particles _ carrying representations @xmath226 .
it could happen that the trivial representation @xmath371 of @xmath255 occurs in ( [ decomp ] ) . in this case , for an observer working only with @xmath255 , _ there will be a non - zero probability of seeing the vacuum_. the irr s of @xmath255 are classified in a similar way as for @xmath13 .
the trivial irr on @xmath372 in the decomposition ( [ decomp ] ) is generated by any state @xmath373 satisfying [ vac ] q^(1)_c & = & _ e , c , + ^(1)_g & = & , [ v2 ] + c^(1 ) & = & .
[ v3 ] we will call such state vacuum . it is not difficult to show that a representation @xmath347,\rho ) $ ] contains states satisfying ( [ vac ] ) if and only if @xmath374 is the identity . furthermore , under the condition @xmath375 all states of @xmath360 , @xmath347,\rho ) $ ] , fulfill equation ( [ vac ] ) .
we are thus left with the conditions ( [ v2 ] ) and ( [ v3 ] ) for defining the vacuum state .
they simply mean that @xmath376 is an identity representation of the group @xmath377 , where @xmath378 is generated by @xmath379 .
note that vacuum decay occurs naturally , for example , in all irr s of @xmath13 of the form @xmath347,\epsilon ) $ ] , where @xmath73 and @xmath136 are in the center of @xmath15 and @xmath380 is the trivial representation of the stability subgroup of @xmath143 , which in this case is the whole of @xmath381 .
the vectors in this representation clearly satisfy all conditions and therefore will decay into vacuum states .
the vector space @xmath360 may contain more than one copy of the identity representation of @xmath377 .
we will denote the set of corresponding orthonormal vectors by @xmath382 finding all @xmath383 in a given decomposition of each @xmath384 is a group theoretical problem that can be solved for specific choices of the discrete group @xmath15 .
we shall not attempt this here .
if @xmath387 , a single geon described by @xmath346 can not decay into the vacuum .
however , two geons colored by @xmath346 and @xmath388 may annihilate each other .
the two geons can shrink to localized objects and at the same time approach each other .
the process can also be interpreted as a change in the scale of observations to long distances . at the end
a point - like object will remain and should be described by the algebra @xmath389 introduced in section [ gasp ] . from the inclusion i^(2):a^(2)_l^(2 ) follows that the space of states @xmath390 of the two geons is a ( reducible ) representation of @xmath391 .
let @xmath226 denote as before the irr s of @xmath285 , with corresponding vector spaces @xmath392 . then _ r_r= _
n^(r , r)_v _ , n^(r , r)_. therefore , the operators of @xmath285 can see the vacuum if @xmath393 is not zero . the vacuum representation and the vacuum probability decay
are given by formulae analogous to ( [ vac])-([prob ] ) .
it is clear now how to describe the decay into the vacuum of an arbitrary number of geons .
consider @xmath12 geons described by representations @xmath394 .
the space of states @xmath395 is a representation of the long distance algebra @xmath396 described in section [ gasp ] .
the system may decay into the vacuum if this representation contains the trivial representation of @xmath287 .
in this work we have developed an algebraic model for topological geons which describes topology change as a purely quantum phenomenon rather than the usual classical sense of cobordisms between two non - homeomorphic spatial manifolds @xmath84 and @xmath397 .
instead , our formalism revealed what an observer , probing the topology of space by using only quantum operators and quantum states , would be able to see .
the key point was that in resorting to a field theory to infer the underlying spatial topology , one could only take into account those operators which were compatible with the scale of observations , since no other operator would a have sensible physical meaning in the theory .
the passage from a larger scale of observations to a smaller one was represented , on a more technical level , by selecting a subalgebra of the original algebra describing the system in the quantum theory .
the quantum states of the system , which in the case of geons give a direct information on the spatial topology , could now decay into the vacuum , leading the would - be observer to conclude that a topology change has ocurred .
there is another , perhaps more intuitive view of the sort of topology change we have envisaged in this paper . as pointed out in section [ s3 ] for the case of vortices , those classical configurations for which holonomies are trivial around some `` topological blob '' , be it a vortex or a geon , are indistinguishable from those in which this `` blob '' is absent , or `` vacuum '' configurations .
if we view quantum states as wave functions , it is clear that their role is to assign a probability to each classical configuration .
a quantum transition to states which are very sharply peaked , or localized at the aforementioned `` vacuum '' configurations will be interpreted by an observer as a quantum topology change .
such states correspond to the vacuum states of section [ s5 ] .
we have restricted ourselves to a simple theory , where complications arising from local degrees of freedom were absent ( the theory we considered is topological in the limit of very low energies ) , and we could concentrate on the topological aspects more unobstrusively . of independent value is the algebra describing the topological geons .
the finite group @xmath15 can be generalized to a lie group @xmath14 .
of special interest is the case @xmath398 , which describes geons in the presence of gravity .
a suitable generalization of our formalism promptly discloses a whole spectrum of geon types in quantum gravity , and many interesting properties of these entities can be explored , as for instance spin - statistics connection .
although this issue has been extensively studied in the literature , our formalism may shed new light on some points .
this subject will be investigated in a forthcoming paper @xcite .
we would like to thank j.c.a .
barata , b.g.c . da cunha , a. momen and s. vaidya for many helpful discussions . the work of a.p .
balachandran was supported by the department of energy , u.s.a .
, under contract number de - fg02 - 85err40231 .
the work of e. batista , i.p .
costa e silva and p. teotonio - sobrinho was supported by fapesp , capes and cnpq respectively .
a.p.b . acknowledges the wonderful hospitality of denjoe oconnor and support of cinvestav at mexico , d.f . while this work was being completed .
see for example : r.p .
geroch , j. math . phys .
* 8 * ( 1967 ) 782 ; + f.j .
tipler , ann . of phys .
* 108 * ( 1977 ) 1 ; + s.w . hawking , nucl .
b144 * ( 1978 ) 349 ; + r. sorkin , phys . rev .
* d33 * ( 1986 ) 978 ; + a. anderson and b. dewitt , found .
* 16 * ( 1986 ) 91 ; + c.a .
manogue , e. copeland and t. dray , parmana * 30 * ( 1988 ) 279 ; + g. horowitz , class .
* 8 * ( 1991 ) 587 .
roberts : `` lectures on algebraic quantum field theory '' , in _ the algebraic theory of superselection sectors .
introduction and recent results _ , edited by d. kastler , world scientific , singapore ( 1990 ) .
de wild propitius and f.a .
bais : _ discrete gauge theories _ proc .
crm - cap summer school `` particles and fields 94 '' , banff , springer verlag ( 1996 ) , hep - th/9511201 + m.d.f . de wild propitius : _ topological interactions in broken gauge theories _ , phd .
thesis university of amsterdam , 1995 , hep - th/9511195 . c. aneziris , a.p .
balachandran , m. bourdeau .
s. jo , t.r .
ramadas and r.d .
sorkin : mod .
. lett . * a4 * ( 1989 ) 331 ; + c. aneziris , a.p .
balachandran , m. boureau , s. jo , r.d . sorkin and t.r .
ramadas : int .
phys . * a4 * ( 1989 ) 5459 . | the topology of orientable @xmath0 spacetimes can be captured by certain lumps of non - trivial topology called topological geons .
they are the topological analogues of conventional solitons .
we give a description of topological geons where the degrees of freedom related to topology are separated from the complete theory that contain metric ( dynamical ) degrees of freedom .
the formalism also allows us to investigate processes of quantum topology change .
they correspond to creation and annihilation of quantum geons .
selection rules for such processes are derived .
= 10000 16 cm = 1ex 0.5 cm 0.5 cm = -0.5 cm -.09 cm = cmss10 = cmss10 at 7pt c i u
# 1| # 1| /#1/#1 |#1 # 1 # 1#1 # 1#1 # 1#1 # 1#1 # 1#1 # 1#1| # 1| # 1 # 1#2#1#2 # 1#1 1.0 cm # 1 |
by nature a fault causes a system to behave abnormally but does not cause the system to shut down @xcite , however if left unattended a fault may lead to a system failure , hence the importance of a fault tolerant control ( ftc ) system . while faults can cause instability in a system , the integration of an ftc scheme significantly increases the ability of the system to maintain overall stability in the presence of a fault @xcite . in this paper
we develop a new active ftc system using an unscented kalman filter ( ukf ) for fault detection and identification ( fdi ) where parameter updates made by the ukf are sent to the nonlinear model predictive control ( nmpc ) based control system . to test the ftc system
the design is applied to the control of a 2d robot model .
+ there are numerous solutions to the ftc design problem with the ftc research community classifying these designs as either _ active _ or _
passive_. passive ftc systems are based on fixed controllers and are designed to be resilient against known faults .
they are designed using robust control techniques for worst case scenarios .
active ftc systems `` actively '' seek the fault and try to gather as much information about it as possible to help the controller overcome any consequential instabilities and are also known as self - repairing , self - designing or fault detection , identification ( diagnosis ) and accommodation schemes .
active schemes are made up of an fdi component and are based on controller redesign , or selection / mixing of pre - designed controllers .
+ the main difficulty with active ftc is on - line reconfiguration which requires detailed information about changes in system parameters .
hence the main role of the fdi subsystem is the gathering of information on parameter changes to assist in controller reconfiguration .
fdi is a key component in an active ftc system and is the most difficult aspect of ftc @xcite . in the early days
most research on fdi was done independently of the controller design and no combined design existed @xcite .
recently there has been some research on the integration of fdi and ftc ; however much remains to be done .
+ techniques commonly used for fdi include artifical intelligence based fdi schemes @xcite , @xcite , @xcite , and multiple model based methods @xcite @xcite where different models are used to describe the dynamics of the system for different operating regimes . sliding mode observers for fault detection have also been used because of their strong robustness to a particular class of uncertainty @xcite .
other methods include analytical fdi methods that are based on estimating the fault through matrix algebra @xcite .
observer based fdi remains the most common approach studied in recent literature where an extended state observer is used to estimate faults .
+ quite often in the design of an ftc system two controllers are utilised ; the main ( or nominal ) controller is designed for the faultless case with the second controller being a compensator designed to handle the faulty case @xcite . in this paper
a number of observer based fdi schemes are investigated and integrated with the nmpc ( nonlinear model predictive control ) controller design of @xcite to develop a full active ftc system . in our design only one controller needs to be designed capable of handling the faultless as well as the fault cases .
the 2d robot model from @xcite is used as a test bed to assist in the development of the ftc system design .
+ this paper is organised as follows ; section [ section : chap4_fdi_tandi ] presents an in - depth look at the fdi methods chosen for investigation , namely the ekf ( extended kalman filter ) , the ukf ( unscented kalman filter ) and the imm ( integrated multiple models ) . an ekf filter , ukf filter , ekf based imm filter and a ukf based imm filter are each designed in section [ section : chap4_probform ] for the purposes of fdi using the 2d robot model of @xcite .
four different active fault tolerant control systems using nmpc ( nonlinear model predictive control ) as the controller design are formulated and implemented in matlab for the 2d robot model .
each of the four different active ftc systems developed are then tested under different conditions in section [ section : chap4_nranda ] .
the tests are designed to evaluate the performance of the filters as well as the interaction of the filter and controller designs .
section [ section : chap4_nranda ] provides a detailed analysis of the test results and concludes with a summary of findings .
based on the findings the ukf filter was chosen as part of the final design of the active ftc system .
section [ subsec : lmpc ] concludes with a comparison between linear mpc and nonlinear mpc as the controller showing that an active ftc system with a nonlinear mpc controller has better overall performance in the event of a failure .
conclusions are given in section [ section : chap4_conclusion ] .
the fault detection schemes considered here are all based on filtering techniques , namely the ekf , the ukf , the ekf based imm and the ukf based imm .
these filters are used to sequentially estimate the state of a dynamic system using a sequence of noisy measurements made on the system .
the state estimates are then utilised to aid in fault detection and control reconfiguration . a general overview and key mathematical concepts are provided for each method below .
the ekf is an extension of the well known kalman filter .
one of the drawbacks of the kalman filter is that it does not provide good estimations for nonlinear systems @xcite .
the ekf approximates ( or linearises ) the nonlinear functions in the state dynamic and measurement models .
there are two main stages during an ekf ( and the general kalman filter ) cycle : predict and update . during the prediction stage
the filter states and covariances are predicted forward one time step as are the measurement predictions . during the update stage corrections
are made to the state predictions via noisy measurements .
a summary of the ekf ( ristic et .
al @xcite ) equations is given below .
the target state @xmath0 and measurement @xmath1 equations propagate according to : @xmath2 where @xmath3 and @xmath4 are random sequences and are mutually independent with zero - mean , white gaussian with covariances @xmath5 and @xmath6 respectively .
the ekf is based on the assumption that local linearisation of the above equations may be a sufficient description of nonlinearity .
the mean and covariance of the underlying gaussian density are computed recursively in a two stage process ( ristic et .
al @xcite ) : + stage 1 : prediction @xmath7 stage 2 : update / correction @xmath8 where @xmath9 , @xmath10 and @xmath11 .
the parameter @xmath12 is commonly known as the kalman gain and @xmath13 is referred to as the innovation covariance .
the innovation @xmath14 is the error between the predicted measurement and the actual measurement of the system .
the matrices @xmath15 and @xmath16 are the local linearisation of the nonlinear functions @xmath17 and @xmath18 . the two matrices are defined as jacobians evaluated at @xmath19 and @xmath20 respectively ( ristic et .
al @xcite ) .
the non - gaussianity of the true posterior density is more evident , for example becomes bimodal or heavily skewed , when the model is highly nonlinear . in this event
the performance of the ekf will significantly degrade .
the ukf addresses the issue of non - gaussianity .
the ukf is a part of a family of nonlinear filters , referred to as linear regression kalman filters , that are based on statistical linearisation rather than analytical linearisation .
the key concept behind these filters is to perform nonlinear filtering using a gaussian representation of the posterior @xmath21 through a set of deterministically chosen sample points .
the true mean and covariance of the gaussian density are completely captured by these sample points up to the second order of nonlinearity , with errors introduced in the third and higher order when propagated through a nonlinear transform .
the ekf on the other hand is only of first order with errors introduced in the second and higher orders .
the filters belonging to this family differ only by the method used to select the sample points i.e. their number , weights and values in the filtering equations are identical and are given below .
the ukf uses an unscented transform for the selection of points in an ekf framework ( ristic et .
al @xcite ) . + we assume that at time @xmath22 the posterior is gaussian : @xmath23 .
the very first step is representing this density via a set of @xmath24 sample points @xmath25 and their weights @xmath26 .
the ukf uses the unscented transform @xcite to select the sample points @xmath25 and weights @xmath27 .
the prediction step is as follows : @xmath28\left[\mathbf{f}_{k-1}\left(\mathcal{x}_{k-1}^i\right ) - \mathbf{\hat{x}}_{k\vert k -1}\right]^\top.\end{aligned}\ ] ] a set of @xmath24 sample points : @xmath29 are used to represent the predicted density : @xmath30 and the predicted measurement becomes : @xmath31 the update step is defined as : @xmath32 where : @xmath33 as can be seen from the above filter equations there is no explicit calculation of jacobians .
consequently these filters can be utilised even when the nonlinear functions @xmath34 and @xmath35 have discontinuities .
the imm belongs to a class of filters called the gaussian sum filters . the main concept here is the approximation of the required posterior density @xmath36 by a gaussian mixture ( ristic et .
al @xcite ) : @xmath37 where @xmath38 are weights that are normalised , @xmath39 .
gaussian sum filters such as the imm are ideal when the posterior density is multimodal because for multimodal densities there is a performance degradation in both the ekf and ukf . at time @xmath40 the state estimate is calculated for each possible current model using @xmath41 filters , with each filter using a different combination of the previous model - conditioned estimates called mixed initial condition .
the algorithm as outlined in bar - shalom et .
al @xcite is : * calculation of the mixing probabilities .
the probability that mode @xmath42 was in effect at time @xmath22 given that @xmath43 is in effect at @xmath40 conditioned on @xmath44 is given by : @xmath45 the above can be written as : @xmath46 where the normalising constants are : @xmath47 * mixing . the mixed initial condition for the filter matched to @xmath48 is calculated starting with @xmath49 : @xmath50 and the corresponding covariance is given by : @xmath51\\ & \cdot \left[\hat{x}^i\left(k-1 \vert k-1\right ) -
\hat{x}^{0}\left(k-1\vert k-1\right)\right]^\intercal\bigg\},\quad\quad i , j = 1,\hdots , r . \end{split}\ ] ] * mode - matched filtering .
the estimates of the states and covariances calculated in step 2 above are used as inputs to the filter matched to @xmath48 which uses @xmath52 to determine @xmath53 and @xmath54 .
the likelihood functions associated to the @xmath41 filters : @xmath55,\ ] ] are calculated using the mixed initial condition and covariance from step 2 : @xmath56,\quad\quad j = 1,\hdots , r,\ ] ] that is : @xmath57,s^j\left[k;p^{0j}\left(k-1\vert k-1\right)\right]\right],\quad\quad j = 1,\hdots , r.\ ] ] * mode probability update .
the mode probabilities are then updated via : @xmath58 where @xmath59 is the normalisation constant and is given by @xmath60 * estimations and covariance combination .
the output is obtained by combining the model - conditioned estimates and covariances : @xmath61\left[\hat{x}^j\left(k\vert k\right ) - \hat{x}\left(k\vert k\right)\right]^\intercal\right\rbrace.\end{aligned}\ ] ] this section outlined the details of the the methods chosen for further investigation , the ekf , the ukf and the imm filters .
these techniques are applied to the 2d robot model in the next section .
to test the different filtering techniques in an fdi context the robot model of @xcite is used for development and testing purposes and forms the plant that is to be controlled as well as the process model of the nmpc controller ( see figure [ fig : chap3_robot eom ] ) .
the equations for this robot model are : @xmath62 where @xmath63 is the @xmath63-coordinate of the point @xmath64 , @xmath65 is the @xmath65-coordinate of the point @xmath64 , @xmath66 is the heading angle , @xmath67 is the right wheel angular velocity , @xmath68 is the left wheel angular velocity and @xmath69 is the speed given by @xmath70 .
+ the nmpc controller is based on the design given in @xcite which solves the following optimal control problem using pseudospectral discretisation : @xmath71 subject to @xmath72 full details of the design are given in @xcite . fifty coincidence points ( @xmath73 )
are used along with a prediction window length of 5 secs ( @xmath74 ) .
the weights @xmath75 and @xmath76 are diagonal matrices with the diagonal values set to 10 , 1 and 1 respectively and found through trial and error . + the fault , to be simulated and tested for , is a punctured wheel .
if a wheel is punctured the radius of the wheel will decrease and so the filters are set up to estimate the radius of the wheel .
four different filters have been designed , the ekf , the ukf , the ekf imm and the ukf imm .
+ the robot parameters used for all simulations are : right wheel radius , @xmath77 , left wheel radius , @xmath78 , distance between wheels , @xmath79 , speed demand is @xmath80 and the input constraints on @xmath67 and @xmath68 are @xmath81 .
the filters are updated at 100hz while the controller is updated at 10hz .
all work was developed using matlab with snopt as the nonlinear programme ( nlp ) solver .
the following subsections detail the design of each of the filters .
the state vector for the ekf consists of the following states : @xmath82^\intercal,\ ] ] where @xmath63 , @xmath65 and @xmath66 are the robot states and @xmath83 and @xmath84 are the right wheel and left wheel radii respectively .
the measurements are assumed to be of the speed , @xmath69 , of the robot : @xmath85 where @xmath86 is additive white noise .
the initial state vector and initial covariance matrix are : @xmath87^\intercal,\quad\quad p(0 ) = \begin{bmatrix } ( 0.5)^2 & 0 & 0 & 0 & 0 \\ 0 & ( 0.5)^2 & 0 & 0 & 0 \\ 0 & 0 & ( 1\pi/180)^2 & 0 & 0 \\ 0 & 0 & 0 & ( 0.5)^2 & 0 \\ 0 & 0 & 0 & 0 & ( 0.5)^2 \end{bmatrix}.\ ] ] the @xmath88 and @xmath89 noise matrices were chosen to be : @xmath90 where @xmath91 is the update rate of the filter .
for the prediction cycle an euler integration scheme is used to predict the states of the ekf forward .
the predicted measurement @xmath92 is given by : @xmath93 given the above information the kalman filter equations given in section [ subsection : chap4_ekf_equations ] are applied to estimate the radius of each wheel in the experiments conducted in section [ section : chap4_nranda ] .
the general structure of the ekf and ukf are very similar in that they both have a prediction and update cycle and produce a single state vector and a corresponding covariance matrix . for the robot model
the state vector is the same as the one given in equation .
the initial state vector , initial covariance matrix , the process noise matrix @xmath88 and the noise covariance matrix @xmath89 all remain the same as those given in subsection [ subsec : chap4_ekf_equations_sims ] .
the ukf algorithm given in subsection [ subsec : chap4_ukf_equations ] is applied to the robot model with @xmath94 @xcite .
the interacting multiple model method , as the name suggests , is made up of multiple models where each model tests a different hypothesis .
four different models ( the terms mode and model are used interchangeably and have the same meaning in the context of imms ) have been designed where : * no fault case . *
@xmath95 right wheel deflation , left wheel no fault . * right wheel no fault , @xmath95 left wheel deflation .
* @xmath95 right wheel deflation , @xmath95 left wheel deflation . during step 3 of the imm algorithm given in section [ subsection : chap4_imm_equations ] a filter such as the ekf is used to update the states and covariances , and both an ekf based imm filter and a ukf based imm filter have been designed .
the initial covariance matrix for each filter and each mode are the same as equation .
the @xmath88 and @xmath89 matrices are those given in equation and the initial state vectors for each filter and mode are : @xmath96^\intercal,\\ \mathbf{x}_2(0 ) & = & \left[x_0,\,\,\,y_0\,\,\,\psi_0\,\,\,1\,\,\,2\right]^\intercal,\\ \mathbf{x}_3(0 ) & = & \left[x_0,\,\,\,y_0\,\,\,\psi_0\,\,\,2\,\,\,1\right]^\intercal,\\ \mathbf{x}_4(0 ) & = & \left[x_0,\,\,\,y_0\,\,\,\psi_0\,\,\,1\,\,\,1\right]^\intercal.\end{aligned}\ ] ] the mixing probabilities or mode probabilities are initially set to : @xmath97^\intercal,\ ] ] and the mode transition probabilities matrix @xmath98 is set to : @xmath99
the robot is required to follow a circular path for all experiments . to simulate the measurement additive white noise
is added to the speed of the robot which is calculated as a part of the truth simulations of the robot movement .
to test the filters four different test cases were set up and each test case was run twice . during the first run the fdi feedback loop is not closed and the filters are used for estimation only . the fdi loop is closed during the second run to investigate the behaviour of the full active ftc controller .
the test cases are as follows : + 1 . no fault .
the objective is to investigate how well the filters estimate the radii of the tyres in a no fault situation .
2 . left wheel @xmath95 puncture . in this case
a puncture is simulated to occur 10 secs into the simulation .
the wheel is assumed to deflate to @xmath95 of its original value instantaneously .
3 . left and right wheel puncture . in this test case
a left wheel puncture is simulated 5 secs into the run and a right wheel puncture is simulated to occur 10 secs into the simulation .
both punctures are assumed to cause an instantaneous reduction of the respective wheel radius to @xmath95 of the original wheel radius .
4 . left wheel linear puncture . in this test case
once again the left wheel is punctured 10 secs into the run however this time the puncture is assumed to follow a linear reduction in wheel radius according to @xmath100 , where @xmath84 represents the left wheel radius reduced from its original value of 2 m down to 0.1 m at a rate of 0.1 m/s and t is the current time . the radius does not drop to 0 m as this caused a complete system failure .
+ the results for each filter are presented in the next four subsections : plots of the speed innovation were produced ( but have been omitted due to space constraints ) where the innovations were plotted along with the calculated @xmath101 uncertainty bounds .
the @xmath101 uncertainty bounds are a @xmath102 confidence interval and the solution ( innovations in this case ) must remain within the bounds @xmath102 of the time . the results for the ekf and ukf filters showed that the speed innovations remained well within the uncertainty bounds throughout the duration of the run both with and without feedback .
the speed innovation plots for the imm filters were also produced .
the results for both the ukf and ekf based imm filters showed that the filters were able to very quickly detect the correct mode of operation .
+ plots of the wheel radius estimates were also produced .
all filters do a very good job of estimating the radius of the tyres with and without feedback .
the imm filters initially have a higher error in the tyre estimate as the estimation is based on a mixture of all the models , however it took only one update for the imm to reach the correct estimate .
+ the wheel speed estimates ( or angular rates ) were also analysed and it was found that the estimates for all four filters were very similar . in the case of no feedback
the estimate was the same as the actual speed ; however in the case where feedback is provided the wheel speeds were quite noisy .
this is a result of calculations based on noisy measurements which is a consequence of feedback .
+ plots of the robot trajectory for all four filters showed that the robot remained on the path with and without feedback which is to be expected in the no fault case .
the speed innovations for all four filters were plotted and the results showed that all filters were able to detect the fault .
the fault occurred at 10 secs into the run , and the plots show that at 10 seconds there was a peak change in the innovation curves .
the corrections / innovations were seen to increase at the time the fault occurred and settle again near zero once the correct estimate was reached .
+ for the imm filters model 3 is the correct match for scenario 2 and both of the imm filters were seen to find the correct mode immediately as mode 3 is the most confident in its estimate .
mode 3 and mode 4 both hypothesise a failure in the left wheel of @xmath95 which is why after the occurrence of the fault the uncertainties do not increase .
however because mode 1 and mode 2 do not hypothesise a fault in the left wheel the uncertainties can be seen to increase once the fault has occurred . the uncertainty in mode 3
was seen to decrease twice as much compared to mode 4 after the fault occurred .
this is because mode 4 hypothesises that both wheels are punctured whereas mode 3 predicts the puncture of only the left wheel .
another point to note is that in the single filter cases feedback did not have much effect on the innovations .
however , in the case of the imm filters the results show that with feedback the filter errors do not grow as rapidly between updates .
the errors are seen to grow very quickly when no feedback is present which is evident in figures [ fig : chap4_scenario2_immukf_velocityerrors_nfb ] and [ fig : chap4_scenario2_immukf_velocityerrors_wfb ] .
uncertainty bounds ( red ) , speed innovations ( blue ) ] uncertainty bounds ( red ) , speed innovations ( blue ) ] the results for the wheel radius estimates showed that the ukf estimates were closer to the actual wheel radius compared to those produced by the ekf . turning on feedback results in the filters reaching a steady estimate faster when compared to the no feedback case .
the imm filters produced slightly better estimates than the single ukf filter and there was very little improvement on the imm estimates compared to the no feedback case . +
the wheel speeds were also analysed and the results for all four filters presented the same trends . without feedback there is much more activity present compared to turning on the feedback .
once the fault occurs the robot yaws to the side with the punctured wheel and demands a faster speed to compensate for the loss in radius . +
the robot trajectories were simulated and all results showed that the robot was only able to remain on the path if feedback from the filter was provided to reconfigure the controller .
although all filter estimates without feedback were excellent , without reconfiguration of the controller the robot could not be made to follow the desired path ( see figure [ fig : chap4_scenario2_traj_ukf ] ) .
the plots of the speed innovations for all filters clearly indicated , from the sudden changes in innovations , the detection of both faults , left wheel at 5 secs and right wheel at 10 secs .
the ekf innovations were found to be consistent , however the innovations produced by the ukf show that , with feedback , the innovation uncertainty begins to grow rapidly between updates whereas without feedback the uncertainty remains constant ( see figure [ fig : chap4_scenario3_ukf_velocityerrors ] ) .
the imm filters show that after 5 secs model 3 is the best model .
however , once the second fault occurs the filters do an excellent job of recognising that mode 4 is the correct match and uncertainties in mode 4 are seen to decrease ( see figure [ fig : chap4_scenario3_immekf_velocityerrors_wfb ] ) .
it was again observed in the no feedback case that the uncertainties on the imm filters grow rapidly between updates and many more corrections are required .
+ uncertainty bounds ( red ) , speed innovations ( blue ) ] uncertainty bounds ( red ) , speed innovations ( blue ) ] the wheel radius estimates produced by the filters showed that the imm filters produce the best estimates of the radii .
the ukf performs slightly better than the ekf , and turning feedback on results in a faster settling time ( see figure [ fig : chap4_scenario3_ukf_wheelradius ] ) .
+ plots of the angular rates achieved by the robot via the ekf showed that once a fault occurred the punctured wheel is required to spin faster in order to compensate for the loss in radius .
the angular rates produced as a result of the ukf , with feedback , resulted in operation at the angular rate constraints .
both imm filters displayed similar behaviour in that once a wheel was punctured it was required to rotate faster to compensate for the loss in radius . + the trajectories produced by each filter for scenario 3 showed that without feedback it is impossible to maintain the robot on the path . reconfiguring the controller on the other hand with estimates from the filters allowed the robot to easily follow the reference path .
an anomaly occurred with the ukf filter where even turning the feedback on did not result in the robot following the path after the occurrence of the second fault .
this could possibly be the result of poor tuning of the filter ( see figure [ fig : chap4_scenario3_traj_ukf ] ) .
+ scenario 4 velocity innovations are presented in figures [ fig : chap4_scenario4_ekf_velocityerrors ] and [ fig : chap4_scenario4_ukf_velocityerrors ] for the ekf and ukf respectively .
the ekf breaks down 5 seconds after the fault occurs , when feedback is turned on as can be seen by the innovations exceeding the covariance bounds . without feedback however the innovations remain within the uncertainty bounds .
the results from the ukf are much better as it produces innovations which remain well within the uncertainty bounds with and without feedback .
both of the imm filters failed , because the fault type of scenario 4 was not modelled as a part of the filter design , i.e. the hypothesis for this type of failure is not accounted for and so no model exists for this failure ( see figure [ fig : chap4_scenario4_immukf_velocityerrors_wfb ] ) .
uncertainty bounds ( red ) , velocity innovations ( blue ) ] uncertainty bounds ( red ) , velocity innovations ( blue ) ] uncertainty bounds ( red ) , velocity innovations ( blue ) ] the wheel radius estimates showed that without a hypothesis on the imm filters the ukf was the only filter able to produce the correct estimates of the wheel radii .
the ukf results clearly indicated that reconfiguring the controller resulted in a faster convergence to the correct estimate .
+ plots of wheel speed produced by all four filters with and without feedback show that with this type of fault both wheels were required to work at their constraints .
+ the robot trajectories as a result of the different filter information were examined and it was found that none of the filters show full compliance with the reference trajectory
. however the ukf was able to maintain the robot on the path the longest .
- this section looks at the behaviour of the imm filters by re - designing the filters to accommodate the fault type covered by scenario 4 .
the filters were modified by adding a fifth model , which hypothesises the left wheel puncturing in the manner described by scenario 4 .
the @xmath88 and @xmath89 matrices remain the same and the initial state vectors for the fifth filter and mode are : @xmath103^\intercal.\end{aligned}\ ] ] the mixing probabilities or mode probabilities become : @xmath104^\intercal,\ ] ] and the mode transition probabilities matrix @xmath98 is redefined as : @xmath105 the speed innovations for the imm ekf are presented in figures [ fig : chap4_scenario4_immekf_velocityerrors_nfb_5models ] and [ fig : chap4_scenario4_immekf_velocityerrors_wfb_5models ] with no feedback and with feedback respectively . as predicted the results show that if a hypothesis is made the imm performs very well .
the ekf as part of imm is able to predict this type of error which it was unable to do as a single filter .
the imm ukf speed innovation plots showed a higher level of confidence in its estimates compared to its ekf counterpart as the uncertainty is lower and consistent . providing feedback in both cases ( ekf and ukf imms )
was shown to increase confidence in the filter estimates .
uncertainty bounds ( red ) , velocity innovations ( blue ) ] uncertainty bounds ( red ) , velocity innovations ( blue ) ] the ukf based imm estimates of the wheel radii for the 5 mode imm are shown in figure [ fig : chap4_scenario4_immukf_wheelradius_5models ] .
the results of both ekf and ukf based imms showed that in both cases the filters do an excellent job of making the correct estimations on the radius of the wheel .
the angular rate plots of the 5 mode imms again showed that the control inputs are required to work at the constraints the majority of the time once the fault occurred .
this result is not dependent on the filter type but rather the fault that has occurred makes it impossible for the robot to achieve the desired task while at the same time respecting its constraints .
+ the trajectory plots showed that for this type of fault where the wheel radius had almost approached zero , the wheels are unable to maintain the reference .
the ukf imm is again able to keep the robot on the path for a longer time than the ekf imm .
the results of the previous section reveal that each of the filters considered exhibits excellent qualities for fault detection and identification .
figure [ fig : radiicomparisons ] shows a comparison of the different filters in terms of wheel radius estimation .
the plot shows the radius estimates for scenario 4 which is the worst case scenario out of the 4 scenarios considered .
the output of the ekf and the 4 mode imm filters have been omitted from the plots as the errors were very high and as a consequence the remaining results were invisible .
the plot shows that in terms of wheel radius estimate performance the ukf performed equally as well as the 5 mode imm filters .
the imm filters were able to reach the correct estimate faster than the single ukf particularly when a sudden change is present as is evident at the 10 second mark .
the ekf performance also dramatically increases when used within an imm configuration .
the drawback of the imm filter is that all possible scenarios must be accounted for .
the method used in this research illustrated the concept of the imm , by showing easy adaptability to different situations , and the speed with which it is able to identify and reach the correct estimate .
however predicting exactly how a fault will occur ( in this case how a tyre will puncture ) is impractical . a more practical implementation
would have been to develop a number of filters each with different process noises that could adapt to all different situations .
the number of filters and the process noise values would have to be determined by trial and error . in any case
an imm performs well if and only if it is equipped to make a hypothesis on the current situation . if the given situation is unaccounted for the filter breaks down . in terms of the single filter , in general the ukf displayed better performance than the ekf , especially in the case of scenario 4 where the non - linearities of the fault caused the single ekf to breakdown . for these reasons
the ukf has been chosen for the final ftc system design . as a result of the findings given in subsection [ subsection : chap4_fandc ]
only the ukf with feedback is implemented to compare nonlinear mpc with linear mpc .
the results given in the next two sections are for scenarios 2 and 4 respectively however plots for only scenario 2 are given due to space constraints .
the velocity innovation plots in figure [ fig : chap4_scenario2_ukf_velocityerrors_linearmpc ] produced by the linear mpc controller show that the innovations remain well within the uncertainty bounds and are approximately zero .
however the uncertainty was seen to be double that produced by the nonlinear controller .
uncertainty bounds ( red ) , speed innovations ( blue ) ] the wheel radii plots give in figure [ fig : chap4_scenario2_ukf_wheelradius_linearmpc ] show that the estimations produced by a nonlinear controller are the same as those produced by the linear controller .
hence the filter performs well even with linear mpc .
the angular rate plots for the linear mpc controller ( figure [ fig : chap4_scenario2_ar_ukf_linearmpc ] ) show that five seconds after the fault occurred the linear controller pushes the wheels to operate at their constraints and is unable to tolerate the faulty condition .
the angular rates produced by the nmpc controller ( figure [ fig : chap4_scenario2_ar_ukf_nmpc ] ) show that it was able to easily adapt to the fault .
this is further illustrated in the trajectory plot given in figure [ fig : chap4_scenario2_traj_ukf_linearmpc ] which clearly shows that the nonlinear mpc controller does an excellent job of keeping the robot on the path despite a faulty wheel whereas the solution produced by the linear controller has diverged .
a point to note here is that while the trajectory tracking is good , the switching behaviour observed in wheel rotation rates is highly undesirable , and in the real world would produce wheel slippage , and high levels of wear on tyres and mechanicals .
the wheel rates produced by the nonlinear mpc controller show a dramatic decrease in this switching behaviour .
it is , however , still present ( figure [ fig : chap4_scenario2_ar_ukf_nmpc ] ) .
this limit cycle behaviour is a negative characteristic that needs to be addressed before this technique can be applied to real systems .
this exercise was purely for proof of concept and is not a practical application .
the model is entirely kinematic and for it to represent a more practical scenario further work is required to eliminate the limit cycling behaviour ; for example adding actuation activity to the cost function and an angular acceleration term to avoid wheel slippage issues .
the speed innovations were plotted for scenario 4 and showed similar trends to those above in that the innovations are quite small ; however the uncertainties with the linear controller are higher than those produced as a result of nonlinear mpc .
the estimates of the wheel radii however are very good and were seen to be the same for both linear and nonlinear controllers . +
as expected the linear controller was unable to maintain the robot on the path .
neither of the controllers were able to drive the robot to the path as the wheel radius had almost reached 0 m making it infeasible for the robot to continue .
the angular rates showed that the linear controller constantly demanded operation at the constraints oscillating between the upper and lower limits continuously .
the nonlinear mpc controller results for wheel angular rates showed that the right wheel oscillates between the upper and lower bounds , however the left wheel is required to constantly work at the upper bound .
the analysis from this work has proven the feasibility of the nmpc controller design with filter estimates for controller reconfiguration as a viable solution to fault tolerant control .
four different filters were compared , the ekf , the ukf , the imm ekf and the imm ukf filters .
the results showed that in terms of fault detection performance the ukf is the best candidate in a trajectory tracking scenario .
+ comparisons were also made between the performance of nonlinear mpc and linear mpc .
the results clearly show that for the purposes of reconfigurable fault tolerant control the nonlinear mpc controller has better performance .
+ the next phase of this research is to implement the nmpc pseudospectral controller with a ukf based fdi subsystem to an aircraft , for fault tolerant flight control .
9 khan , r. , williams , p. , riseborough p. , rao , a. and hill , r. , _ designing a nonlinear model predictive controller for fault tolerant flight control_. arxiv e - prints 2016 ; http://adsabs.harvard.edu/abs/2016arxiv160901529k , [ last accessed : 07 - 09 - 2016 ] .
cork lr , walker r , dunn s. _ fault detection , identification and accommodation techniques for unmanned airborne vehicles_. australian international aerospace congress ( aiac ) 2005 ; http://eprints.qut.edu.au/1729/ , [ last accessed : 07 - 01 - 2016 ] .
boskovic jd ,
li sm , mehra rk .
_ on - line failure detection and identification ( fdi ) and adaptive reconfigurable control ( arc ) in aerospace applications_. proceedings of the 2001 american control conference 2001 ; 4:26252626 .
jiong - sang y , bin j , jian lw , yan l. _ discrete - time actuator fault estimation design for flight control application_. 7th international conference on control , automation , robotics and vision 2002 ; 3:1287 - 1292 .
doi:10.1109/icarcv.2002.1234958 .
wang w , hameed t , ren z. _ extended state observer - based robust fault - tolerant controller for flight control surface failures_. icemi09 .
9th international conference on electronic measurement & instruments 2009 ; 3610 .
wan ea , van der merwe r. _ the unscented kalman filter for nonlinear estimation_. the ieee 2000 adaptive systems for signal processing , communications , and control symposium 2000 ; 153 - 158 . doi:10.1109/asspcc.2000.882463 | this paper develops a new active fault tolerant control system based on the concept of analytical redundancy .
the novel design consists of an observation filter based fault detection and identification system integrated with a nonlinear model predictive controller .
a number of observation filters were designed , integrated with the nonlinear controller and tested before reaching the final design which comprises an unscented kalman filter for fault detection and identification together with a nonlinear model predictive controller to form an active fault tolerant control system design .
keywords : fault tolerant control , fault detection and identification , model predictive control , nonlinear control , observation filters . |
most , if not all , minor planet populations are host to multiple systems , including binaries and trinaries ( see walsh 2009 for review ) . these multiples have a vast array of properties , from extremely short - period , contact binaries , to systems that are exceedingly widely separated and have mutual periods of many years . some have tiny satellites , thought to be collisional fragments blown off of their parent , and others have components of near - equal size and mass .
the characteristics of these systems represent a treasure trove of information about the properties of the objects that compose them , the environment they are embedded in , and the dynamical history of their parent population . through their orbital separations and periods , binaries offer the only way to measure the mass of these distant objects , which when combined with radius measurements determine these objects bulk densities which in turn provide information about composition and physical structure ( such as porosity ) . in the last 10 years
, trans - neptunian populations have been found to host to a very high fraction of binary systems .
the binary fraction varies in sub - populations from @xmath8% in the low - inclination component of the classical kuiper belt to just a few percent in other dynamical classes ( noll et al . 2008a ) . given the low interaction rates of the kuiper belt populations today , forming
such a large number of binary systems has proven a theoretical challenge , especially with the limited information available for the components of these systems ( eg .
, goldreich et al .
2002 , weidenschilling 2002 , noll et al .
2008b , schlichting & sari 2008a & b ) .
trans - neptunian binaries ( tnbs ) are distinguished from binary systems elsewhere in the solar system by the high frequency of near - equal sized binaries , and by the presence of binaries with extremely wide separations and long mutual - orbit periods . widely - separated , long - period tnbs are difficult to create and very sensitive to perturbation ( nesvorn et al .
2011 , parker & kavelaars 2010 , petit & mousis 2004 ) , and make valuable tracers of the dynamical and collisional conditions over the history of the outer solar system .
the orbital , compositional and statistical properties of these binaries constrain the total mass and dynamical history of the various populations , with important implications for theories of solar system formation and evolution .
there are several dozen known tnbs , but only a small subset has measured orbital parameters ( noll et al .
2008b , naoz et al .
2010 , grundy et al .
2011 ) . most of these , in turn , are relatively tightly - bound binaries that have been characterized by observations from space ( systems with mutual semi - major axis much less than 5% of their hill radius , eg . , grundy et al .
2009 & 2011 ) .
two tnb systems with moderately widely - separated components have published mutual orbits ( 1998 ww@xmath9 and teharonhiawako / sawiskera , both with mutual semi - major axes of order 5% of their hill radius ) , but the widest tnbs ( those with mutual semi - major axes substantially exceeding 5% of their hill sphere ) have not been well - characterized to date , with a preliminary orbit estimate available only for the system 2001 qw@xmath2 ( petit et al .
2008 ) . such wide - separation , near - equal mass binaries all have low heliocentric inclinations , indicating that they belong to the cold component of the classical kuiper belt .
these wide binaries make up at least 1.5% of the known cold classical belt objects ( lin et al .
2010 ) .
the creation of large - separation , near - equal mass binaries is most likely during the formation phase of tnos , as most proposed formation scenarios require a much higher space density of objects than is observed today .
additionally , most tnbs have identically - colored components , with differences between primary and secondary color being much smaller than the large color variation seen across the entire population of tnos ( benecchi et al .
this suggests that each component formed from similar material in a similar region of a locally homogeneous protoplanetary disk with global variations in composition .
different mechanisms proposed for binary formation dominate under different dynamical conditions ( eg .
, schlichting & sari 2008a ) . if the dynamical properties of the systems today can be taken to be representative of their primordial distribution , they can probe the dynamical conditions of the primordial kuiper belt during the formation phase .
however , any intervening violent dynamical events , such as collisions ( petit & mousis 2004 , nesvorn et al .
2011 ) or close encounters with giant planets ( parker & kavelaars 2010 ) can leave today s mutual orbit distribution substantially altered from its original state .
it is critical to measure the orbital properties of a large sample of tnbs , as well as perform dynamical studies of possible sources of orbital modification , in order to understand the full extent of information about the formation and history of the outer solar system encoded in these systems .
we have collected astrometric measurements of a sample of seven of the widest - known tnbs for an extended period , covering four to nine years of orbital motion for each system .
these observations have allowed us to compute accurate mutual orbits for our sample of ultra - wide tnbs , and from these orbits we derive system mass and a host of other characteristics . in the first part of this paper
, we outline the nomenclature we adopt to describe these systems and their host populations ( 1.1 ) , our sample selection criteria ( 2 ) , details of our observational campaign and data reduction techniques ( 3 ) , and mutual orbit fitting algorithm ( 4 ) . in the latter part
, we describe the mutual orbit fits ( 5 ) and compare them to the properties of other binary populations , and derive geometric albedos for each system given reasonable assumptions of bulk density ( 6 ) .
finally , we conclude with a discussion of possible formation mechanisms and implications for the early history of the outer solar system , susceptibility of these systems to disruption by collisions and neptune scattering , and present future surveys abilities to discover and characterize a large sample of these ultra - wide tnbs . in this paper ,
we compare several sub - populations of trans - neptunian objects and their various orbital properties . in order to facilitate a clear understanding of the nomenclature we use to describe these populations and their properties , we provide an outline here .
a binary s mutual orbit properties will be described either as a `` mutual '' property or denoted by the subscript `` @xmath10 . '' in contrast , the properties of the orbit of the binary s barycenter around the sun will be described as an `` outer '' property or denoted by the subscript `` out . '' in order to compare the properties of our sample with those in the literature , some dynamical classification is required .
we adhere roughly to the gladman et al .
( 2008 ) nomenclature when discussing outer orbit properties . in this paper
, we frequently deal with binaries that belong to the following dynamical classes : * `` classical : '' non - resonant objects in the range 34 au@xmath11 au , 37 au @xmath12 au . * `` cold classical : '' subset of `` classical '' objects with low orbital excitations and confined in semi - major axis . when dividing samples , we assign `` classical '' binaries with @xmath13 , @xmath14 au , and 42.4 @xmath15 au to this population .
referred to as cc population in the text .
* `` hot classical : '' subset of `` classical '' objects with higher mean orbital excitations , and an extension to lower pericenter than the cc population .
when dividing samples , we assign `` classical '' binaries with @xmath16 , @xmath17 au , @xmath18 or @xmath19 au to this population .
referred to as hc population in the text .
this dynamical classification is somewhat different from that adopted by grundy et al .
( 2011 ) , and several binaries in that work which were classified as `` extended scattered '' fall into our hc classification .
in addition , we compare the outer orbital distributions of binary sub - samples with the canada - france ecliptic plane survey ( cfeps ) l7 synthetic model of the kuiper belt . we compare the cc binary sub - sample with the composite of the `` stirred '' and `` kernel '' sub - components of the synthetic kuiper belt model , and refer to the composite of these sub - components as cc - l7 .
we compare the hc binary sub - sample with the `` hot '' sub - component of the synthetic kuiper belt model , and refer to this sub - component as hc - l7 . in reality , any simple inclination cut is insufficient to determine which population a given object truly belongs to , as both the cc and hc populations overlap significantly .
according to the cfeps l7 model , most relatively bright objects below @xmath20 of inclination actually belong to the hc - l7 population .
we stress that while we will refer to a given object as a `` cc '' binary or an `` hc '' binary , there is no way to absolutely verify the parent population for a given single object .
however , we show later that the sub - samples of binaries which fall into our cc and hc classifications have dynamically distinct mutual orbit distributions , and the outer orbit distributions of cc binaries suggest that they are in fact members of the cc - l7 population , and likewise the hc binaries outer orbit distribution is consistent with the hc - l7 distribution .
lccccccccc + + name & date range & @xmath21 & @xmath22 & @xmath23 & @xmath24 & @xmath25@xmath26 & + & & & & ( @xmath27 ) & ( @xmath27 ) & ( @xmath27 ) & @xmath28 ( au ) & @xmath29 & @xmath30 ( @xmath31 ) + 2000 cf@xmath4 & 2002 - 2011 & 12 & 50 & 23.85@xmath32 & 0.72(5 ) & 7.70 & 43.84 & 0.0362 & 0.528 + 2001 qw@xmath2 & 2001 - 2010 & 35 & 88 & 23.16@xmath33 & 0.03(5 ) & 7.51 & 43.98 & 0.0242 & 4.808 + 2003 un@xmath34 & 2003 - 2010 & 14 & 60 & 22.7@xmath35 & 0.88(6 ) & 7.5 & 42.62 & 0.0035 & 3.069 + 2005 eo@xmath36 & 2005 - 2011 & 12 & 52 & 22.45@xmath37 & 1.45(3 ) & 6.59 & 45.62 & 0.0679 & 3.415 + 2006 br@xmath34 & 2006 - 2011 & 20 & 66 & 23.0@xmath32 & 0.50(4 ) & 7.3 & 43.80 & 0.0393 & 1.157 + 2006
jz@xmath38 & 2006 - 2011 & 15 & 56 & 22.7@xmath32 & 0.98(2 ) & 6.9 & 44.70 & 0.0804 & 3.550 + 2006 ch@xmath0 & 2004 - 2010 & 15 & 47 & 23.0@xmath32 & 0.44(5 ) & 7.0 & 45.74 & 0.0362 & 1.791 +
since we seek to characterize the widest binaries ( which have correspondingly long mutual periods ) , we opted to pursue a ground - based observation campaign .
we chose our sample based on the following criteria : 1 .
the system had no well - characterized orbit in literature .
the separation at discovery was @xmath39 .
the magnitude difference between the system s primary and secondary was less than 1.7 , indicating a near - equal mass system ( @xmath40 ) . at the time of our sample selection , there were seven systems that met these criteria : 2000 cf@xmath4 , 2001 qw@xmath2 , 2003 un@xmath34 , 2005 eo@xmath36 , and three objects discovered over the course of cfeps , with internal designations b7qa4 , l5c02 , and heav .
the binary nature of 2000 cf@xmath4 was presented in noll et al .
( 2002 ) , while the binary natures of 2003 un@xmath34 and 2005 eo@xmath36 were presented in kern ( 2006 ) .
provisional orbital characterization for 2001 qw@xmath2 was presented in petit et al .
the cfeps target l5c02 was identified as binary by lin et al .
( 2010 ) , and the binary nature of b7qa4 and heav are presented here for the first time .
all of the outer orbits of this sample of objects fall into our cc classification , and have very low outer inclinations and eccentricities . two other cfeps targets , l4q10 and l4k12 , were initially included in our sample , due to data from the canada - france - hawaii telescope ( cfht ) suggesting that they were elongated in a manner consistent with a near - equal mass binary with a separation of order @xmath41 . however , follow - up observations in very good seeing did not bear out their putative binary nature , and they were removed from our target list .
the three cfeps objects in our sample also have mpc designations : b7qa4 is 2006 br@xmath34 , heav is 2006 jz@xmath38 , and l5c02 is 2006 ch@xmath0 . throughout this paper
we will refer to these three systems via their mpc titles .
a targeted observational campaign from 20082011 was executed from _
gemini north _ using the _ gemini multi - object spectrograph _ in imaging mode , taken with the @xmath42 filter .
observations were queue scheduled , with stringent requirements on image quality ( frequently at the expense of photometric conditions ) . by requiring modest visit times ( @xmath8 minutes ) , excellent seeing could be obtained without the use of adaptive optics , in some cases with full - width half - max of @xmath43 or better .
additional observations during this period were made from _ vlt _ with the _ fors2 _ instrument , though image quality requirements were not held to the same stringent limits .
single - epoch observations were also made in april 2010 from _ magellan _ with the _ megacam _ imager .
significant archival data also exists for all systems .
we used the _ solar system object search _
gwyn 2011 , in prep ) service provided by the _ _ canadian astronomy data centre _ _ to locate and download images from the _ cfht _ and _ hst _ public archives that contained our targets , and we also located images of our targets from the _ mayall _ , _ hale _ , and _ wiyn _ telescopes . literature astrometric measurements are available for some systems as well . the relative astrometry for 2001 qw@xmath2 published in petit et al .
( 2008 ) is also included in our fit for that system .
astrometric measurements of 2003 un@xmath34 and 2005 eo@xmath36 were presented in kern ( 2006 ) , and we include those measurements in our fits for these systems . combining all these data sources , the smallest number of observed epochs for any binary in our sample is 12 visits for 2005 eo@xmath36 , while the largest number is 35 visits for 2001 qw@xmath2 . during most visits ,
more than one usable image was acquired .
the number of visits and total number of images from which astrometric measurements were made are listed in table 1 .
astrometric solutions were generated for each individual image , matched to the j2000 coordinate system using reference stars in the _ usno b _ astrometric catalog .
whenever possible , the catalog stellar positions were corrected for proper motion since their last observed epoch , and uncertainties in the final reference positions reflected the original astrometric precision and the integrated uncertainty due to the stated uncertainty in proper motion over the intervening time .
the brightest 100 non - saturated stars were identified in the ccd on which the binary was located ( in the case of multi - chip imagers , other chips in the array were not used to constrain the astrometric solution ) , and their ( @xmath44 ) positions ( and uncertainties ) extracted using _ sextractor _
( bertin & arnouts , 1996 ) . in the case of an image with a good initial world
coordinate system ( wcs ) estimate ( eg . , _ gemini gmos _ images ) , this wcs was used to estimate each stars ra and dec position in the j2000 system and the nearest neighbor in the _ usno b _ catalog was identified as its matching counterpart .
if an image did not have a good initial wcs , a robust pattern - recognition algorithm identified probable rough corrections to the wcs , applied these corrections , and then identified nearest - neighbor stars in the _ usno b _ catalog .
once _ usno b _
( ra , dec ) positions were matched to ( @xmath44 ) positions in the image , the _ iraf _ package _ ccmap _ was used to generate a wcs solution .
because this package does not handle positional uncertainties in either the image or reference positions , the positions of each matched star is cloned 1,000 times , adding gaussian noise to each position consistent with the ( ra , dec ) and ( @xmath44 ) uncertainties .
iterative fitting followed by automatic clipping of outliers in these thousands of cloned sources allowed _ ccmap _ to generate a robust astrometric solution automatically which reflected the uncertainties in the absolute and measured positions of the reference stars .
this allowed a more robust automatic solution to be derived with little input from the operator for each image processed .
after the first pass of _ ccmap _ , more matches are searched for in the image with the _ usno b _ catalog reference stars , and upon flagging any new matches the _ ccmap _ routine is called again .
this matching and wcs - fitting process is iterated ten times for every frame . .
bottom left : synthetic psf model of binary components .
top right : image residuals after subtracting binary and other point - sources in the image .
bottom right : relative contributions of both psf components .
same stretch is applied to all images , and flux scaling is linear.,scaledwidth=50.0% ] the lowest - order astrometric solution merited by the distortions of the optics of each imager was used in each case . in the case of _ gemini gmos _ images , this was a simple rotation and fixed pixel scale . in the case of most other imagers , the distortion across the area of a single ccd was low enough such that the addition of independent @xmath45 and @xmath46 pixel scales , as well as a skew term , was sufficient . in the case of hst observations , the internal hst astrometric solutions and distortion corrections were used .
once an astrometric solution had been found for an image , the relative astrometry of the binary pair in that image was then extracted using a custom point - spread function ( psf ) fitting routine .
the psf model we adopted was a sum of two elliptical gaussian components , with arbitrary long - axis orientation for each component .
the wider of the two gaussians ( the psf wings ) is arbitrarily limited to contain less than @xmath47 the flux of the narrower `` core '' gaussian to prevent runaway solutions with extremely wide wings . adopting a non - circular psf
was required due to the fact that most images were obtained with sidereal tracking rates , and over the period of integration the psf of the binary s components became somewhat stretched along their direction of motion .
a variant of the same algorithm used for fitting the mutual binary orbits ( described in the following section ) was used to minimize the residuals in a sky - subtracted 40@xmath4840 pixel region centered on the binary .
an initial interactive step is used to identify all the point sources in this region and flag the two associated with the binary .
initial estimates for the amplitude and @xmath49 of the psf are made automatically , and these values are fed into a markov - chain monte carlo algorithm which finds the psf model and array of point source positions and amplitudes that produces the smallest residuals . because the image model varies with the number of point sources in the 40@xmath4840 pixel region , the minimum number of free parameters the algorithm must search over is 9 ( two point sources , and a two - component psf model forced to be circular ) while the maximum number of free parameters ever treated was 22 ( five point sources and a two - component psf model with arbitrary rotation and ellipticity for each component ) .
example of data from the _ gemini _ observatory is illustrated in figures 1 and 2 , along with the image residuals after psf fitting and subtraction , and the model of the binary system .
each fit is visually inspected , and in general we found that our adopted psf model produced extremely low residuals . in some cases where the two binary components were blended
, we performed a check to verify that the extra degrees of freedom added by allowing the psf to be elongated was not skewing the measured astrometry . in these cases , we performed a second , independent fit using a circular psf and compared the measured astrometry for the binary components .
in general , we found excellent agreement between the two . in cases of very strong blending ,
fits with the elongated psf would occasionally have trouble converging to a stable solution , and in these cases we adopted the values from the circular psf fits .
we did not attempt to determine any upper limits on separation based on completely unresolved images , and therefore at present such images do not contribute to our mutual orbit fits .
relative astrometry was recorded as separation ( in arcseconds ) and position angle ( in degrees east of north ) , and the observation date was taken at the central julian date of each observation .
uncertainty in the relative astrometry was estimated as @xmath50 , where @xmath51 indicates signal - to - noise ratio of primary or secondary .
a noise floor is set by the uncertainty in the astrometric solution for each frame .
these psf fits also returned relative photometry for each system , and the mean @xmath52-mag measured in the _ gemini gmos _
@xmath42 filter during well - resolved visits is also included in table 1 .
relatively few observations were made in photometric conditions , as image quality was our primary concern .
all mpc targets in our sample have published absolute system photometry in various bands , but only 2001 qw@xmath2 and 2005 eo@xmath36 have @xmath27-band magnitudes in literature . 2000
cf@xmath4 , 2006 br@xmath34 , 2006 jz@xmath38 , and 2006 ch@xmath0 were all imaged on photometric nights from cfht , and _ elixir _ processed images were used to determine @xmath27-band system magnitudes for these systems . the @xmath27-band magnitude of 2003 un@xmath34 was determined from observations on a single night from gemini north , though the absolute calibration of these particular images is poor and the resulting photometric uncertainty is relatively large .
the basic operations performed by our mutual - orbit fitting routine are , given an initial guess of mutual orbit properties , to solve kepler s equation in order to determine the relative system geometry at the time of observation ( accounting for the light - time delay between the system and observer ) , then rotate the system in space to account for its orientation with respect to the ecliptic .
finally , the code `` observes '' the system by applying a second rotation to account for the variation in viewing geometry induced by the relative motion between the earth and the binary , and projects the result onto the sky plane given the separation between the observer and the system .
to fit our observations , we chose to adopt the metropolis algorithm @xmath53 minimization routine ( metropolis et al . 1953 ) , using a similar implementation to that described in simard et al .
( 2002 ) , who utilized the metropolis algorithm to fit a 12 dimensional bulge + disk model to images of galaxies .
this algorithm is robust to complicated topology in parameter space , and can be easily adjusted to thoroughly explore parameter space at the expense of speed .
the metropolis algorithm is a markov - chain monte carlo technique which , after an initial burn - in period , occasionally makes `` bad '' decisions , allowing it to diffuse out of local minima .
after a number of iterations , the choice of new parameter values can be informed by previous values , improving the speed of convergence in complicated parameter - space topology .
a binary mutual orbit has seven free parameters , and in our implementation we chose these to be the following : mutual semi - major axis ( @xmath54 ) , eccentricity ( @xmath55 ) , period ( @xmath56 ) , mean anomaly ( @xmath57 , valid at a defined jd ) , inclination ( @xmath58 ) , longitude of the ascending node ( @xmath59 ) , and argument of pericenter ( @xmath60 ) .
the last three angular parameters are defined with respect to the j2000 ecliptic . for nearly circular orbits ,
@xmath57 and @xmath61 become degenerate and alternate choice of basis is preferred ; however , we were not presented with a circumstance where altering the basis used in our code was merited , as the binaries we observed all have significant eccentricity . for all orbit
fits , 15 metropolis algorithm threads are run simultaneously , and each compare their final best - fit value and sampled orbit space to identify the global best - fit and statistically acceptable range of parameters after an additional test of the error distribution .
ccccc + + name & jd & sep & pa & err + & ( d ) & ( @xmath62 ) & ( @xmath63 ) & ( ) + 2001qw322&2452117.92296&3.6056&290.0270&0.2030 + 2001qw322&2452145.74085&3.8915&291.3190&0.1722 + 2001qw322&2452145.76181&3.8683&287.1030&0.1845 + 2001qw322&2452145.78289&3.9600&290.0990&0.1168 + 2001qw322&2452146.75689&3.7377&292.2420&0.2255 + table 2 is published in its entirety in the electronic + edition of the _
astrophysical journal_. a portion is + shown here for guidance regarding its form and content . because our estimates of the astrometric uncertainty of each observation may not reflect their true distribution , ( i.e.
, we do nt know the properties of the distribution of errors on the individual measurements _ a priori _ ) , we perform a test to bootstrap the 68% and 95% confidence intervals for the @xmath53 distribution .
this test simply assumes that the best - fit orbit is the `` truth , '' and estimates the observed error distribution around the best - fit .
the test randomly draws @xmath64 measurements of the observed error from the best - fit orbit from the pool of @xmath64 real observations of a given target ( sampled with replacement , so observations may be repeated in the resampled list ) . after this resampling
, we compute a new `` observed '' @xmath65 based off of the resampled list and best - fit orbit .
we store this new @xmath65 statistic and repeat the process 10,000 times , building up a distribution of best - fit @xmath65 values which `` might have been . ''
we find the 68% and 95% upper confidence intervals on this distribution , and set those to be our @xmath53 thresholds for statistically acceptable mutual orbit fits .
orbits that fall below these @xmath53 thresholds are saved , and their distribution is used to generate the uncertainties for each orbital parameter . this analysis has led to extremely well - behaved fitting behavior , with consistently nested uncertainty contours after every addition of new data to a given fit . the mutual orbit fitting code was tested by reproducing the orbital parameters for the pluto - charon system based on synthetic data generated by the _ jpl horizons _ system , and reproducing the orbital parameters of 2001 xr@xmath66 and 2004 pb@xmath67 as published in grundy et al .
( 2009 ) to an accuracy well within their stated uncertainties .
we did not explore the robustness of our method for determining confidence intervals with these tests , as we did not have sufficient numbers of real observations for the two published mutual orbits we tested .
the astrometric measurements for each system and comparison to fit orbits are illustrated in figures [ mpc_binaries ] and [ cfeps_binaries ] .
these figures project each system onto the sky plane in physical separation units , removing the variation in the observed scale of the systems due to the change in the observer - system separation over the course of a year .
an example of the mutual astrometry for all the tnbs used in these figures is shown in table 2 .
the full table is available in the electronic edition .
the best - fit and uncertainty ( given as the extrema of each parameter from the distribution of orbits allowed at the 68% level of confidence ) of all fitted mutual orbit parameters are listed in table 3 .
additionally , table 4 contains derived parameters ; specifically , the system mass @xmath68 , mutual semi - major axis to hill - radius fraction @xmath69 , mutual inclination @xmath70 , mutual argument of pericenter @xmath71 , and mutual pericenter separation in multiples of primary radii @xmath72 .
figure [ polar_contour ] illustrates the @xmath69 , @xmath55 , and @xmath73 fits and the 68% and 95% uncertainties in these parameters for each system .
l r@l r@l c r@l r@l r@l r@l c + + name & & & @xmath55 & & & & & epoch + & & & & & & & + 2000 cf@xmath4 & @xmath74 & @xmath75 & @xmath76 & @xmath77 & @xmath78 & @xmath79 & @xmath80 & @xmath81 & @xmath82 & @xmath83 & @xmath84 & @xmath85 & @xmath84 & 2454880.96 + 2001 qw@xmath2 & @xmath76 & @xmath86 & @xmath87 & @xmath88 & @xmath89 & @xmath90 & @xmath91 & @xmath92 & @xmath93 & @xmath94 & @xmath95 & @xmath96 & @xmath97 & 2452117.92 + 2003 un@xmath34 & @xmath98 & @xmath99 & @xmath100 & @xmath101 & @xmath102 & @xmath103 & @xmath104 & @xmath105 & @xmath106 & @xmath107 & @xmath108 & @xmath109 & @xmath110 & 2452963.77 + 2005 eo@xmath36 & @xmath111 & @xmath112 & @xmath113 & @xmath114 & @xmath115 & @xmath116 & @xmath117 & @xmath118 & @xmath119 & @xmath120 & @xmath121 & @xmath122 & @xmath123 & 2453440.94 + 2006 br@xmath34 & @xmath124 & @xmath125 & @xmath126 & @xmath127 & @xmath128 & @xmath129 & @xmath130 & @xmath131 & @xmath132 & @xmath133 & @xmath134 & @xmath135 & @xmath134 & 2455153.08 + 2006 jz@xmath38 & @xmath74 & @xmath136 & @xmath126 & @xmath137 & @xmath138 & @xmath139 & @xmath140 & @xmath141 & @xmath142 & @xmath143 & @xmath132 & @xmath144 & @xmath145 & 2455007.86 + 2006 ch@xmath0 & @xmath124 & @xmath146 & @xmath74 & @xmath147 & @xmath148 & @xmath149 & @xmath150 & @xmath151 & @xmath152 & @xmath153 & @xmath154 & @xmath155 & @xmath82 & 2455190.06 + l r@l r@l r@l r@l r@l + + name & & & & & + & & & & & & & + 2000 cf@xmath4 & @xmath156 & @xmath157 & @xmath158 & @xmath159 & @xmath79 & @xmath160 & @xmath161 & @xmath84 & @xmath162 & @xmath163 + 2001 qw@xmath2 & @xmath164 & @xmath165 & @xmath158 & @xmath166 & @xmath167 & @xmath168 & @xmath169 & @xmath170 & @xmath171 & @xmath172 + 2003 un@xmath34 & @xmath139 & @xmath173 & @xmath158 & @xmath174 & @xmath175 & @xmath176 & @xmath177 & @xmath178 & @xmath179 & @xmath180 + 2005 eo@xmath36 & @xmath164 & @xmath181 & @xmath158 & @xmath182 & @xmath183 & @xmath184 & @xmath185 & @xmath121 & @xmath186 & @xmath187 + 2006 br@xmath34 & @xmath98 & @xmath188 & @xmath158 & @xmath189 & @xmath190 & @xmath130 & @xmath139 & @xmath191 & @xmath192 & @xmath193 + 2006 jz@xmath38 & @xmath194 & @xmath195 & @xmath158 & @xmath196 & @xmath194 & @xmath197 & @xmath96 & @xmath198 & @xmath199 & @xmath200 + 2006 ch@xmath0 & @xmath100 & @xmath201 & @xmath158 & @xmath202 & @xmath203 & @xmath204 & @xmath205 & @xmath154 & @xmath206 & @xmath207 + @xmath208 : primary radius @xmath209 assumes @xmath210 gram @xmath211 . in the following section
, we describe the derived mutual orbital parameters listed in table 4 and illustrated in figure [ polar_contour ] .
system mass is simply calculated from kepler s laws , @xmath212 while the classical hill radius for a binary in orbit around the sun is defined as @xmath213 where @xmath28 and @xmath29 is are the heliocentric semi - major axis and eccentricity of the binary system s barycenter .
primary radius is found from the system mass by assuming that both components have the same albedo and density , and is given by @xmath214 where @xmath215 is the adopted bulk density and @xmath24 is the magnitude difference between the primary and secondary components of the binary .
mutual inclination is the angle between the pole vector of the binary mutual orbit @xmath216 and that of the outer orbit @xmath217 , and can be found by @xmath218 .
the pole vectors for either orbit can be found by ( eg . , naoz et al .
2010 ) : @xmath219 the mutual argument of pericenter @xmath71 ( critical for estimating the extent of kozai oscillations ) is the angle between the ascending node ( with respect to the outer orbit ) and pericenter in the plane of the mutual orbit
. it can be found by : @xmath220 ) , \ ] ] where @xmath221 and @xmath222 .
systems within a broad range of @xmath73 centered on @xmath223 may be subject to large oscillations in @xmath73 and @xmath55 due to the kozai effect ( kozai 1962 ) . over the oscillation cycles two values
are conserved : one depending on the initial mutual eccentricity and inclination , and the other depending on both these and the mutual argument of pericenter .
following perets & naoz ( 2008 ) , we adopt the following form for these two conserved values : @xmath224 and @xmath225 with some algebraic manipulation and the constraint that eccentricity and inclination minima and maxima occur when @xmath226 or @xmath223 , these two constants determine the maximum and minimum eccentricity and inclination reached by a given binary system during its kozai cycle .
we calculate these values , and the amplitudes of the eccentricity excursions experienced by each binary system is listed in table 5 .
also listed are the predicted minimum pericenter separations ( occurring during the highest eccentricity phase of the kozai cycle ) in multiples of primary radii , as close encounters during kozai cycles may lead to modification of the mutual orbit due to tidal friction ( fabrycky & tremaine 2007 , perets & naoz 2009 , brown et al .
details of systems with large - amplitude kozai oscillations will be discussed on an object - by - object basis in the following sections .
the kozai effect can be easily suppressed by other effects , including permanent asymmetries in the mass distribution of the component bodies ( ragozzine 2009 ) . however , this suppression only occurs for relatively small semi - major axes , and therefore the kozai oscillations of these wide systems are unlikely to be suppressed . with several epochs of hst data , and
nearly nine years of observational baseline ( covering most of a single 11 year mutual orbit period ) , 2000 cf@xmath4 has one of the best - measured orbits in our sample with mutual semi - major axis and period uncertainties at the 1% level .
it is also the lowest - mass system in our sample , and the second - most weakly bound ( behind only 2001 qw@xmath2 ) . because of its low mass ( currently the lowest mass of any known tnb ) , this system has the smallest estimated primary radius of any system in our sample ( assuming all objects share a common density ) , estimated to be @xmath227 km given a bulk density of 1 gram @xmath211 .
2000 cf@xmath4 is not subject to strong kozai oscillations , and its pericenter separation is always greater than @xmath228 primary radii , so there is little chance of mutual tides having modified its mutual orbit .
the pole solution for 2000 cf@xmath4 is non - degenerate at greater than 95% confidence , and the system is retrograde .
its mutual pole vector is only @xmath229 degrees from anti - aligned with its outer orbit s pole vector , making it one of the most pole - parallel systems known . , given the data presented in petit et al .
2008 ( gray contours ) and the complete set of data used in this work ( black contours ) using the fitting algorithm described in the text .
stars mark best - fit orbits .
68% and 95% confidence intervals are marked by the solid and dashed contours , respectively.,scaledwidth=50.0% ] at the outset , our current results for the mutual orbit of 2001 qw@xmath2 appear inconsistent with the orbit published in petit et al .
( 2008 ) . by fitting only the astrometric data used in that paper with our new code
, we find a much larger range of allowable mutual orbit solutions than was presented in the previous study .
this is especially notable for allowing much higher values of mutual eccentricity .
this larger range of allowed orbital parameters completely overlaps with our current orbit fit , as illustrated in figure [ comp_qw322 ] .
we suspect that the difference is due to two factors : our more thorough fitting algorithm and our different statistical analysis of the allowed @xmath53 range . the allowed @xmath53 range in petit et al .
( 2008 ) was much smaller than that determined by the methods used in this work .
the components of this system remain photometrically indistinguishable , with @xmath24 consistent with 0 . when making astrometric measurements , we have arbitrarily assigned the northernmost object in the discovery epoch as the system primary .
observations have been frequent enough that there is no possibility for confusion between system components , based on continuity of the orbital motion .
the observations in our dataset cover approximately nine years , sampling slightly over 50% of the best - fit mutual orbital period of 17.0 years .
because the observations have been frequent and of high quality , this limited sample of the orbital motion of the system is very constraining .
mutual semi - major axis , period , and eccentricity are all known to better than 10% accuracy .
angular and derived parameters are similarly well known .
2001 qw@xmath2 is subject to strong kozai oscillations , with the nominal best - fit orbit implying eccentricity variations between @xmath230 .
its current orbit is therefore near its highest eccentricity phase . in physical units ,
this system remains the most widely separated binary minor planet known , with a mutual semi - major axis of @xmath231 km .
it is also likely the most weakly bound binary minor planet known , with its measured @xmath69 of @xmath232 exceeding the current estimate for the outer satellite of the main - belt asteroid ( 3749 ) balam of @xmath233 ( marchis et al .
several other known main - belt asteroid binaries have estimates of @xmath69 shown in richardson & walsh ( 2006 ) which are similar to or slightly higher than that measured for 2001 qw@xmath2 , but these estimates are based on single - epoch observations and may not reflect the true orbits and masses of these systems . the pole solution for 2001 qw@xmath2
is non - degenerate at greater than 95% confidence , and the system is retrograde with a mutual inclination of @xmath234 .
the orbit of 2003 un@xmath34 is the least - well constrained in our sample .
the fit relies heavily on astrometry published in kern ( 2006 ) for pinning the 20032004 astrometry , and only three data points constrain the 20052008 astrometry .
recent data has proved relatively discriminatory , and the mutual semi - major axis and period are both known to better than 10% accuracy , but the derived system mass is only constrained to @xmath235% accuracy .
the @xmath24 we adopt for 2003 un@xmath34 is determined from two well - resolved visits from gemini north .
kern ( 2006 ) finds a highly variable @xmath24 for this system , suggesting one or both components may have significant lightcurve , and our adopted @xmath24 may not reflect this variability .
this system is likely subject to minor kozai oscillations , but the amplitude of these oscillations are comparable to the uncertainty in the currently measured mutual eccentricity .
the pole solution for 2003 un@xmath34 is non - degenerate at greater than 95% confidence , and the system is prograde with a mutual inclination of @xmath236 .
the orbit of 2005 eo@xmath36 also relies heavily on astrometry published in kern ( 2006 ) , and it has the fewest observed epochs ( 12 ) of any of our systems . nevertheless , the recent measurements from _ vlt _ and _ gemini _ have provided reasonably tight constraints on the orbit properties .
this system has the largest @xmath24 in our sample at 1.45 in @xmath27 .
it also has the largest primary radius ( assuming all objects share a common density ) , which we estimate to be @xmath237 km given a bulk density of 1 gram @xmath211 .
2005 eo@xmath36 is not subject to significant kozai oscillations , and its minimum pericenter separation is at least 692 primary radii .
the pole solution for 2005 eo@xmath36 is non - degenerate at greater than 95% confidence , and the system is prograde with a mutual inclination of @xmath238 , making it one of the lowest mutual inclination systems known .
nearly all observations of 2006 br@xmath34 used in our astrometric fit come from _ gemini _ , and the orbit has been sampled for just over a single orbital period .
its orbit is very well constrained , with mutual semi - major axis and period known to 1% accuracy and mutual eccentricity to better than 4% accuracy .
the pole solution for 2006 br@xmath34 is non - degenerate at greater than 95% confidence , and the system is prograde with a mutual inclination of @xmath239 , making it the most inclined system in our sample .
2006 br@xmath34 is subject to strong kozai oscillations , with the nominal best - fit orbit implying eccentricity variations between @xmath240 .
its current orbit is therefore near the lowest point of its eccentricity cycle .
its pericenter passages are still widely separated ( never lower than 143 primary radii ) and mutual tides are not a concern for this system .
s kozai oscillations .
top panel : representative sample of 107 orbits consistent with 2006 ch@xmath0 astrometry at @xmath241 level , showing their current eccentricity ( @xmath242 ) and the maximum eccentricity ( @xmath243 ) they reach over the course of a kozai cycle , with the best - fit orbit marked by the large point .
bottom panel : same sample of orbits , but now illustrating their best - fit system mass ( @xmath68 ) and the minimum pericenter separation in multiples of primary radii ( @xmath244 , assuming @xmath210 gram @xmath211).,scaledwidth=50.0% ] the observations of 2006 ch@xmath0 cover well over a single orbital period , though as figure [ cfeps_binaries ] illustrates its on - sky behavior is quite complex and well time - sampled observations were necessary to accurately constrain this system s mutual orbital properties .
this system has the highest well - measured mutual eccentricity of any binary minor planet , at @xmath245 .
the orbit of the outermost satellite of the trinary asteroid ( 3749 ) balam has been estimated to rival this at @xmath246 ( marchis et al .
2008 ) , but its value is poorly constrained .
a fascinating consequence of this extreme mutual eccentricity is that over the course of its mutual orbit , the secondary of 2006 ch@xmath0 will subtend an angle ranging from to @xmath247 ( at mutual apocenter ) to @xmath248 ( at mutual pericenter ) as viewed from the surface of the primary from one - fifth to over three times the angular size of the moon on the sky .
2006 ch@xmath0 may be subject to strong kozai oscillations , with the nominal best - fit orbit implying eccentricity variations between @xmath249 .
its current orbit is therefore near its highest eccentricity phase .
the pole solution for 2006 ch@xmath0 is non - degenerate at greater than 95% confidence , and the system is retrograde with a mutual inclination of @xmath250 .
this system is also the most tightly bound in our sample , with @xmath251 .
because of the orbits high eccentricity , the error distributions in @xmath54 and @xmath56 conspire to produce a large uncertainty in the derived system mass , and @xmath68 remains uncertain at the 40% level .
the high mutual eccentricity phases of 2006 ch@xmath0 s kozai cycles lead to very close passages between the primary and secondary . at its current mutual eccentricity ,
pericenter passages occur at @xmath252 primary radii ( again given a bulk density of 1 gram @xmath211 ) , which may be wide enough that mutual tides do not cause orbital modification . however , during the high eccentricity phase of its kozai cycle , the pericenter separation of 2006 ch@xmath0 drops much lower to @xmath253 .
figure [ l5c02_kozai ] illustrates the distribution of mutual pericenter separation versus system mass .
if we argue that 2006 ch@xmath0 must have survived roughly in its current orbital configuration for the age of the solar system , then we would prefer higher minimum pericenter separations to keep the system from experiencing tidal evolution . from figure [ l5c02_kozai ]
we see that orbit fits which have higher pericenter separations are lower - mass solutions
. however , the tidal evolution of highly eccentric , highly inclined binary systems is poorly understood , and future work is merited to determine if limiting the tidal evolution of 2006 ch@xmath0 would provide useful priors for further constraining its mutual orbit .
the observations of 2006 jz@xmath38 span slightly more than a single period , and figure [ cfeps_binaries ] illustrates that like 2006 ch@xmath0 the on - sky behavior of this system is also complex .
this system is also very highly eccentric , at @xmath254 , making it the second - highest eccentricity tnb known behind 2006 ch@xmath0 .
its mutual semi - major axis remains somewhat poorly constrained at 16% uncertainty , while the mutual period is known to better than 5% uncertainty .
the derived system mass remains highly uncertain for the same reason as 2006 ch@xmath0 , and uncertainties in @xmath68 remain at the 40% level .
the pole solution of 2006 jz@xmath38 is non - degenerate at the 95% level , and the system is prograde with the lowest mutual inclination of any known tnb at just @xmath255 . due to this low inclination ,
2006 jz@xmath38 is not subject to notable kozai oscillations . despite its high eccentricity and relatively large primary ( @xmath256 km ) , mutual pericenter passages are always at least 64 primary radii , making this system much less susceptible to possible mutual tidal effects than 2006 ch@xmath0 . comparing the tnbs studied here to previously characterized tnbs
, we see that they are much more widely separated in terms of their hill sphere occupation .
figure [ polar_contour ] illustrates the mutual semi - major axis to hill radius fraction for the ultra - wide tnbs characterized in this work and the systems listed in grundy et al .
( 2011 ) ; the widest known tnb with a well - characterized orbit not in our sample is teharonhiawako / sawiskera at @xmath257 .
grundy et al . ( 2011 ) showed that the previously known widely - separated , loosely - bound tnbs are only found on dynamically cold heliocentric orbits . here
we seek to confirm this relationship , and identify which dynamical population hosts the wide binaries ; is it the cold classical kuiper belt , or can a low - inclination extension of the hot classical kuiper belt plausibly host the widely - separated binary systems ?
figure [ helio_excite ] illustrates the outer orbital excitation ( given by @xmath258 ) versus @xmath69 . as in grundy
et al . ( 2011 ) , we confirm that only dynamically cold populations host wide binaries .
further , systems with pericenters which suggest current or past encounters with neptune have relatively low @xmath69 , supporting the destructive nature of such encounters presented by parker & kavelaars ( 2010 ) . to quantify the difference between the binaries found in dynamically cold and hot populations , we compare the @xmath69 distributions of binaries falling into our cc classification with all other binaries .
we find that these two samples are inconsistent with being drawn from the same @xmath69 distribution , with the ks test rejecting this hypothesis at greater than 99.9% confidence .
therefore , the binaries which fall into our crude dynamical classification of the cold classical kuiper belt have distinctly different characteristics than binaries hosted in other populations . to further clarify the host population of the wide binaries , we compare the outer orbital excitation distribution of the binaries with the distribution of the cc - l7 and hc - l7 orbital excitations .
the distribution of orbital excitations for these binary samples and the l7 model populations are illustrated in figure [ helio_hist ] . since
most binaries considered here have been discovered in near - ecliptic surveys , there is significant bias against the detection of high inclination binaries .
the bias is less significant when considering low inclination populations , so comparisons between the biased cc binary sample with the de - biased cc - l7 model population are reasonably fair . comparing the cc binary sample to just the low inclination objects in the de - biased hc - l7 model population is also meaningful , as observational biases are not significant for the low - inclination extension of the hc - l7 population .
we use the ks test to determine if we can reject the following two hypotheses : that the cc binaries are drawn from the cc - l7 orbital excitation distribution , or that the cc binaries are drawn from the excitation distribution of low - inclination ( @xmath259 ) hc - l7 orbits .
additionally , we determine whether just the seven ultra - wide binaries characterized in this work ( with no outer orbit constraints placed on them ) could be drawn from the cc - l7 or @xmath259 hc - l7 excitation distributions , and whether _ all _ binaries with @xmath260 and @xmath261 could be drawn from the same excitation distributions .
we rule out ( @xmath262 ) that the cc binaries are drawn from any subset of the hc - l7 excitation distribution , while it is plausible ( ks test can not reject at high confidence ) that they are drawn from the cc - l7 excitation distribution .
we also find that , without any prior cuts on heliocentric orbits , neither the seven ultra - wide binaries characterized in this work nor all binaries with @xmath260 and @xmath261 can be drawn from the low - inclination hc - l7 excitation distribution ( @xmath262 in both cases ) , while both can be plausibly drawn from the cc - l7 excitation distribution .
this further confirms that wide binaries are intimately linked to the dynamically cold population of the classical kuiper belt it can not be that they are predominantly hosted by a low - inclination extension of the hot classical kuiper belt .
lll r + + name & @xmath263 & @xmath264 & @xmath265@xmath208 + 2001 qw@xmath2 & @xmath266 & @xmath267 & @xmath268 + 2003 un@xmath34 & @xmath269 & @xmath270 & @xmath271 + 2006 br@xmath34 & @xmath272 & @xmath273 & @xmath274 + 2006 ch@xmath0 & @xmath275 & @xmath276 & @xmath253 + 2000 cf@xmath4 & @xmath78 & @xmath277 & @xmath278 + 2005 eo@xmath36 & @xmath279 & @xmath280 & @xmath281 + 2006 jz@xmath38 & @xmath282 & @xmath283 & @xmath284 + + @xmath208 : primary radius @xmath209 assumes @xmath210 gram @xmath211 .
the mutual inclination of a tnb system is one of the most challenging parameters to measure , as there is a mirror degeneracy in the pole solution which can only be broken after sufficient time has elapsed for the the observer s viewing geometry of the binary system has changed enough to discern the system s true orientation .
however , the distribution of mutual inclinations and the ratio of prograde to retrograde orbits holds significant implications for formation scenarios ( schlichting & sari 2008b , noll et al .
2008b ) and for the ongoing evolution of the binary orbit ( fabrycky & tremaine 2007 , perets & naoz 2009 ) .
as illustrated in figure [ polar_contour ] , all seven systems characterized in this work now have non - degenerate pole solutions : four prograde and three retrograde .
the prograde - to - retrograde ratio and its 95% poisson counting uncertainty for the ultra - wide tnbs is therefore @xmath285 .
if we fold in the systems with non - degenerate pole solutions presented in this work and those in grundy et al .
( 2011 ) which fall into our cc sub - sample and meet our near - equal mass criteria , we find that the ensemble prograde - to - retrograde ratio for dynamically cold , near - equal mass tnbs is @xmath286 . @xmath223 .
dashed black line shows ultra - wide binaries inclinations when smoothed over entire kozai cycle of each object .
gray dashed line : uniform distribution @xmath287 .
grey dash - dotted line : uniform distribution modified by removing any orbits with kozai cycles which drive @xmath288 as described in the text .
gray solid line : inclination distribution of form @xmath289.,scaledwidth=50.0% ] the distribution of the mutual orbit inclinations of the ultra - wide binaries is illustrated in figure [ inc_dist ] , along with the inclination distribution of tightly - bound tnbs in literature , taken from the orbits compiled by grundy et al .
( 2011 ) . due to the relative lack of high - inclination and excess of low - inclination wide binaries ,
we find it highly unlikely that the wide binaries inclinations are drawn from the same inclination distribution as the tighter tnbs , with the kolmogorov - smirnov ( ks ) test rejecting this hypothesis at @xmath290% confidence .
additionally , we find that the tight binaries in literature are consistent with having their inclinations drawn from a uniform distribution ( @xmath291 , with the ks test rejecting this hypothesis at only @xmath292% confidence ) , while the ultra - wide tnbs inclinations are inconsistent with being drawn from the same uniform distribution ( ks test rejecting this hypothesis at @xmath293% confidence ) . a simple way to understand the strength of this rejection is to note that in a uniform inclination distribution , 50% of the inclinations are @xmath294 , while there are no systems in our sample with inclinations so high . since the probability of randomly drawing an inclination @xmath295 is 0.5 each time , after sampling seven systems the probability of every system having inclination @xmath295
is the same as landing seven coin flips head - up in a row , @xmath296 .
when restricting the tightly - bound literature binaries to only those in the cc sample , the significance of the rejection of the wide binaries being drawn from the same distribution drops to only @xmath297% however , the dynamically cold binaries from literature remain consistent with a uniform distribution ( ks test rejecting this hypothesis at only 40% confidence ) .
it should be noted that observational bias works against the detection of low inclination systems ( with respect to the ecliptic ) , as they present edge - on geometry and their average projected separations are lower than high inclination systems .
thus , our relative lack of high inclination objects is not due to observational bias . when considering all the inclinations the wide binaries reach during their kozai cycle , we find that a @xmath298 times a gaussian distribution ( frequently used to describe the inclination distribution of heliocentric kuiper belt orbits , eg . , brown 2001 ) centered at @xmath299 with width @xmath300 is compellingly similar to the kozai - modified distribution .
we find that widths between @xmath301 are consistent with the observed ( non - kozai ) distribution within the 95% confidence interval .
the difference in inclination distributions between both populations could either be due to cosmogonic variations inherent in different formation mechanisms , or due to evolutionary processes .
fabrycky & tremaine ( 2007 ) showed that tidal friction coupled with kozai oscillations can cause wide , high - inclination systems to shrink and become circularized , creating a paucity of widely - separated , high - inclination binaries , and an excess of tight binaries near the critical inclinations for kozai cycles ( @xmath302 and @xmath303 ) .
this is referred to as the kozai cycles with tidal friction ( kctf ) mechanism , and its plausibility as a significant evolutionary mechanism for binary minor planets was confirmed by ragozzine ( 2009 ) and perets & naoz ( 2009 ) . while small numbers of objects and other complicating effects may prohibit the detection of an increase of tight binaries near the critical inclinations , the observed paucity of high - inclination , ultra - wide tnbs is suggestive of the kctf effect in action .
we have performed a cursory test to compare the observed inclination distribution to the outcomes of kctf .
we determine the maximum eccentricity @xmath243 each binary system can reach for a grid of initial inclinations @xmath304 and arguments of pericenter @xmath305 , assuming they begin with initial @xmath55 and @xmath54 equal to their present values .
we then determine the fraction of phase space as a function of @xmath304 ( assuming @xmath304 was initially uniformly distributed on a sphere ) which do not lead to an @xmath243 which cause each system s pericenter to drop below a critical number of primary radii in other words , we require that @xmath306 . ragozzine ( 2009 ) showed in full numerical simulations with reasonable assumptions that tidal dissipation became significant at @xmath307 for the binary system orcus / vanth , which is more massive than the binaries studied here .
however , the binary system 2006 ch@xmath0 has kozai cycles which take it to @xmath308 , and presumably its existence suggests that such pericenter separations are stable for a significant fraction of the age of the solar system . as such , we chose the limit @xmath309 for our cursory kctf test , and compare the resulting `` kctf - modified '' inclination distributions to the observed ultra - wide tnb inclination distribution ( illustrated in figure [ inc_dist ] ) .
we find that the resulting distribution ( and any with smaller @xmath310 cutoff ) is ruled out at @xmath311% confidence .
we conclude that while the kctf mechanism may have modified the orbits of some very high - inclination ultra - wide tnbs , it alone is not enough to explain the current mutual inclination distribution of these systems . to verify this ,
more comprehensive studies of the effects of kctf on systems like those presented here are needed .
if kctf is not sufficient to explain the mutual inclination distribution of the ultra - wide tnbs , then we must consider cosmogonic effects .
the primordial poles of the wide binaries may have preferred orientations orthogonal to the ecliptic plane , suggesting formation in a very cold disk ( noll et al .
it should be noted that if the inclination distribution of the ultra - wide tnbs is non - uniform , and the binaries are subjected to small perturbations over their lifetimes ( e.g. , collisions ) , then the mutual inclination distribution will always tend to become _ more _ random over time , approaching a uniform distribution .
as such , the primordial inclination distribution would have to have favored low inclinations even more strongly than the current distribution does .
we explore the effects of collisions on the mutual inclination distribution in an upcoming paper .
gram @xmath211 , while top / right axes show the values assuming @xmath312 gram @xmath211 .
circles mark primary radii , triangles connected with dashed line mark secondary radii .
black points represent binaries characterized in this work , gray points represent values derived from literature for the low - inclination classical tnbs 1998 ww@xmath9 , borasisi / pabu , logos / zoe and teharonhiawako / sawiskera . dotted lines represent contours of constant _ h _ magnitude .
gray region marks @xmath313 , likely unpopulated due to flux limits of current binary searches.,scaledwidth=50.0% ] given the dynamically - derived masses and visible photometry for each system , we can explore the albedos and densities for the component bodies in these binaries . without radiometric measurements to ascertain the albedo independently ,
the density and albedo remain degenerate . however , by assuming physically plausible values for the component densities
, we can estimate the implied albedos for each object ; the results of this exercise are illustrated in figure [ albedo ] and listed in table 6 .
generally , albedos for our sample of ultra - wide tnbs are found to be consistent with those measured radiometrically for larger solitary cold classical kuiper belt objects ( eg . , brucker et al .
2009 ) , and range from 9%30% , assuming @xmath314 gram @xmath211 ( 6.4%21% assuming @xmath315 gram @xmath211 ) .
figure [ albedo ] also includes estimates of the albedos and radii of four other binaries from literature , and these systems were selected as members of the cc sample which had estimates of their @xmath6-band magnitude .
the albedos of these four literature systems range from 5.4%28% ( with @xmath314 gram @xmath211 ) .
lll r + + + name & @xmath316 & @xmath317 & note + & ( @xmath27 ) & ( km ) & + 2000 cf@xmath4 & 0.30@xmath318 & 31.8@xmath319 & @xmath320 + 2001 qw@xmath2 & 0.093@xmath321 & 64@xmath191&@xmath320 + 2003 un@xmath34 & 0.09@xmath322 & 62@xmath323&@xmath320 + 2005 eo@xmath36 & 0.15@xmath324 & 76.2@xmath325&@xmath320 + 2006 br@xmath34 & 0.22@xmath324 & 44.9@xmath326&@xmath320 + 2006 jz@xmath38 & 0.17@xmath327 & 61@xmath328&@xmath320 + 2006 ch@xmath0 & 0.23@xmath329 & 50@xmath330&@xmath320 + 1998 ww@xmath9 & 0.054@xmath331 & 74@xmath84 & @xmath332 + teharonhiawako & 0.147@xmath333 & 76.0@xmath334 & @xmath335 + borasisi & 0.192@xmath336 & 81.6@xmath337 & @xmath338 + logos & 0.28@xmath339 & 41.1@xmath337 & @xmath340 + + @xmath320 : adopting @xmath341 .
@xmath342 : adopting @xmath343 .
@xmath344 : photometry from veillet et al .
@xmath345 : photometry from benecchi et al .
@xmath346 : photometry from delsanti et al .
@xmath347 : photometry from jewitt & luu ( 2001 ) .
@xmath348 : mass from grundy et al .
( 2011 ) .
the apparent trend of increasing albedo with decreasing radius visible in figure [ albedo ] is due to a selection effect : in a flux - limited survey , any physically small object detected must have a high albedo .
the apparent trend here is consistent with a flux limit somewhat less than @xmath349 .
we note that , in general , the observations which discovered the primary of a given binary system were not the observations which discovered the binary nature of the system .
thus , this flux limit seems to apply to the primary absolute magnitude , and the fact that secondary absolute magnitudes scatter across the @xmath350 line reflects deeper follow - up observations identifying the secondaries .
the lack of a strong group of small , high - albedo binaries along with the lack of large , high - albedo binaries suggest that relatively low albedos may be more common than high albedos in the cold classical kuiper belt .
all seven binaries in figure [ albedo ] with primary radius greater than 55 km have nominal albedos less than 20% , while all four binaries with primary radius smaller than 55 km have nominal albedos greater than 20% .
given the steepness of the size distribution of low - inclination tnos in this size range ( @xmath351 , fraser & kavelaars 2009 ) , we would expect there to be roughly nine times as many objects in the radius radius range from 3055 km than at all radii larger than 55 km . since the sample contains only roughly 0.57 times as many binaries in the smaller size range , we posit that the rest of the expected small binaries are missing due to having low albedos , making them invisible to the flux limits of the surveys which discovered the binary systems in our sample .
we adopt an ansatz albedo distribution of a gaussian centered at @xmath352 , and clipped such that @xmath353 ( comparable to the lowest measured albedo in our sample ) : @xmath354 additionally , we adopt a size distribution with slope @xmath355 , and estimate the flux limit at discovery to be @xmath350 .
when we draw a radius from this size distribution , we assign it an albedo from our gaussian albedo distribution and determine if it is bright enough to have been observed by our synthetic survey ( brighter than @xmath350 ) .
we then compare the properties of these synthetic `` observed '' systems with the systems which were actually observed ; we vary the width @xmath356 of the albedo distribution until the ks test can rule out that either the distribution of real radii or the distribution of real albedos are drawn from the synthetic `` observed '' distributions .
we find that at 95% confidence , the observed range of real albedos implies that the albedo distribution must have a width @xmath357 or there would be too few high albedo detections .
additionally , given the observed real primary radius distribution , the albedo distribution must have a width @xmath358 or there would be too many small objects discovered .
figure [ albedo_dist ] illustrates the distribution of albedos and primary radii of the real binaries in comparison to the synthetic `` observed '' distributions of these parameters
. the lack of a strong group of small , high - albedo binaries may also be explained in the case of a more uniform albedo distribution by positing that the binary fraction decreases drastically for decreasing radii , similar to the prediction of nesvorn et al .
. however , with the addition of a varying binary fraction with size as a new degree of freedom , the current sample size is not sufficient to quantitatively constrain such behavior at this time .
we note that given the large range of albedos observed in this population , any sharp features in the trend of binary fraction with radius will be blurred if considering only the absolute magnitude of systems ( as done in nesvorn et al .
2011 ) , and any such features will be much more evident when radii derived from mass measurements or radiometric measurements are used in place of absolute magnitudes .
since the discovery of the first tnbs , a number of possible formation pathways have been posited . in the following discussion , we consider those mechanisms most likely to form widely - separated , near - equal mass systems , and compare the predicted outcomes of each of these pathways to our observed sample . originally described in goldreich et al .
( 2002 ) , these mechanisms were further investigated by noll et al .
( 2008b ) and schlichting & sari ( 2008 a & b ) .
the @xmath359 mechanism posits that binaries are captured when two passing solitary objects can disperse some excess kinetic energy into a sea of smaller bodies and become bound , while the @xmath360 mechanism instead sends the excess kinetic energy away through scattering a third large body .
schlichting & sari ( 2008b ) show that models like @xmath359 , which rely on a smooth dissipation process to capture binaries , will dominate the binary formation rate only in conditions where the relative velocity between planetesimals @xmath361 is much less than the hill velocity , @xmath362 , where @xmath363 is the heliocentric orbital period .
they also show that under these conditions , the binary mutual inclinations will be dominantly retrograde , predicting a prograde - to - retrograde ratio @xmath364 .
the measured wide binary inclinations exclude such an extreme ratio of prograde to retrograde systems .
therefore , we can rule out this mechanism for forming wide binaries , unless an intervening dynamical process can be invoked to re - orient a large number of binary systems or preferentially destroy retrograde binary systems .
we estimate that starting from a primordial prograde - to - retrograde ratio of 0.03 , at least 22% of wide binary systems would have to be re - oriented in order to not be ruled out at greater than 95% confidence by the current observed prograde - to - retrograde ratio . under more energetic conditions , where the relative velocity between planetesimals exceeds @xmath365 , schlichting & sari ( 2008b )
show that three - body interactions ( @xmath360 models ) will dominate the binary formation rate . in this regime
, they find that roughly equal numbers of prograde and retrograde systems are formed , consistent with the observed distribution of inclinations .
however , they also show that only systems with separations of order @xmath366 tend to survive the formation phase , and binary formation rates drop dramatically as @xmath361 increases . as our observed binary systems have separations of order
@xmath367 , the velocity dispersion in the primordial disk could not have exceeded 24 times @xmath365 if this formation mechanism applied
. additionally , noll et al . ( 2008b ) suggest that formation in a dynamically cold disk ( @xmath368 ) should produce aligned orbit poles .
since the wide binaries seem to prefer low mutual inclinations , this argues for formation in a dynamically cold disk , but not so cold as to allow @xmath359 to dominate and produce a large fraction of retrograde systems .
together , the widely - separated components , lack of clear preference for retrograde orbits , but apparent preference for low mutual inclinations all point toward the velocity dispersion being approximately equal to the hill velocity . this represents a fine - tuning problem ( noll et al .
2008b ) , for there is no clear _ a priori _ reason to expect that @xmath369 . additionally , it is not clear whether the balance between the @xmath360 and @xmath359 mechanisms at @xmath369 would simultaneously produce widely separated binaries , aligned poles , _ and _ roughly equal numbers of prograde and retrograde orbits .
funato et al .
( 2004 ) suggest that multiple exchange reactions ( where one object in a binary system is swapped for a passerby ) can produce very widely separated binary systems .
however , systems as widely separated as those in our sample formed through exchange reactions all have very high eccentricities ( @xmath370 , see figure [ polar_contour ] ) .
the systems 2006 ch@xmath0 and 2006 jz@xmath38 have present eccentricity values consistent with these predictions , but all other systems are presently inconsistent with such high eccentricities .
several systems are subject to large oscillations of their inclination and eccentricity due to kozai cycles , but these oscillations do not carry them to eccentricities as high as predicted by exchange reactions . thus , it seems unlikely that this mechanism dominated binary formation .
astakhov et al .
( 2005 ) simulated the effect of chaotic transient binaries on stable binary formation .
they found that two objects temporarily caught in their mutual chaotic layer could become stabilized by dynamical friction due to a sea of small objects effectively adding an enhancement to the @xmath359 mechanism due to transient , chaotic orbits .
this mechanism is referred to as chaos - assisted capture .
they find mutual eccentricities spanning the range of those observed in our sample , but the separations they find for binaries formed by chaos - assisted capture do not extend to as high as those found for the ultra - wide tnbs ( see figure [ polar_contour ] ) .
additionally , schlichting & sari ( 2008a ) argue that the enhancement due to these transient captures is not significant , and that formation should proceed as they found for the @xmath359 and @xmath360 mechanisms .
we conclude that it is unlikely that this mechanism dominated the ultra - wide tnb formation rate .
recently , another mechanism has been proposed to form kuiper belt binaries .
operating with the framework of planetesimal formation through rapid gravitational collapse in a turbulent disk , nesvorn et al .
( 2010 ) suggest that binaries may form as a cloud of cm - scale particles collapses and fragments .
this model produces binaries very efficiently , and their properties can vary widely .
mass ratios of order unity are produced , and semi - major axes from @xmath371 km are produced for systems with primary radii ranging from tens to hundreds of kilometers .
broad ranges of inclination and eccentricity can be produced for all semi - major axes .
additionally , this mechanism has the attractive feature of producing a natural explanation for the correlated colors of binary components ( benecchi et al .
2009 ) , in contrast to the broad range of colors exhibited between different binary systems . for cc tnbs ( solid histogram ) and all other tnbs with known orbits ( dashed histogram ) .
also shown is results of gravitational collapse binary formation simulations by nesvorn et al .
( 2010 ) , selecting simulations with @xmath372 ( initial clump rotation ) , @xmath373 ( cross - section modifier ) , and final mass ratio @xmath374 10.,scaledwidth=50.0% ] we compare the results of a subset of these simulations to our observed mutual orbits .
we select only those simulations which produce binaries with final mass ratio @xmath374 10 , and with initial particle - swarm rotation @xmath375 ( where @xmath376 , and @xmath57 and @xmath377 are the total initial mass and radius of the swarm ) and collisional cross - section enhancements ( to account for the lower resolution of the simulation compared to reality ) @xmath378 .
see nesvorn et al .
( 2010 ) for a more thorough description of these parameters and their importance .
figure [ polar_contour ] illustrates the mutual eccentricity and @xmath69 for all the synthetic orbits formed by this mechanism which meets these criteria , and in general we find that they appear to mimic the distribution of real orbits surprisingly well .
figure [ arh_hist ] illustrates the distribution of @xmath69 for the synthetic orbits compared to the cc binaries and all the other binaries .
we find that the @xmath69 distribution of the cc binaries is statistically indistinguishable from the synthetic orbit distribution generated by gravitational collapse , while we rule out that the other binary populations were drawn from the same distribution as the synthetic orbits at a high level of confidence .
this agreement is encouraging , but it remains preliminary .
the initial conditions for binary formation by gravitational collapse is still highly uncertain , and it remains unclear if this mechanism can produce a large number of retrograde binaries or if the binaries it produces will exhibit a preference for low mutual inclinations .
additionally , while the @xmath69 distributions appear to agree , there are likely a number of biases in the current observed distribution and any conclusions drawn from it must be taken with caution .
the binary fraction is measured to be @xmath8% in the cold classical kuiper belt ( noll et al .
2008a ) at the limit of hst resolution , while the fraction of ultra - wide systems is approximately 1.5% ( lin et al .
2010 ) ; therefore , the ultra - wide binaries make up approximately 5% of the current cc binary population .
however , comparing the systems with measured orbits , we see that the ultra - wide systems are over - represented , with our sample alone making up @xmath379% of all the cc binaries with measured mutual orbits . taking this into account , it appears that the nesvorn et al .
( 2010 ) model actually over - produces wide binaries compared to the present population .
there is additional concern since the current @xmath69 distribution may not represent the primordial distribution ; petit & mousis ( 2004 ) showed that these wide binary systems were likely much more common in the past , and may have been reduced in number due to collisional disruption .
we discuss this problem in more detail in the following section . because of their low binding energy , ultra - wide tnbs are very susceptible to disruption by a variety of perturbations .
parker & kavelaars ( 2010 ) simulated the transplantation of the cold classical kuiper belt from closer to the sun via neptune scattering as suggested by levison et al .
( 2008 ) , and found that wide binaries are very efficiently destroyed by close encounters with neptune before they can be implanted in the current kuiper belt .
the binaries we characterize in this work are all wide enough to be easily stripped by this process , and comparing the @xmath69 and eccentricity of our systems to the results of parker & kavelaars ( 2010 ) , we find destruction probabilities ranging from at least 75% for 2006 br@xmath34 to over 98% for 2001 qw@xmath2 . furthermore , such interactions would likely have randomized the orbit poles of the surviving wide binaries , not leaving behind the aligned poles we see today . taken together
, we estimate that if our sample represents a population which has been subjected to disruption by neptune scattering , the initial population of ultra - wide tnbs would have to have been roughly 13 times larger than the current population . if we take the lower limit of the current wide binary fraction of the cold classical kuiper belt estimated by lin et al .
( 2010 ) at 1.5% , this indicates that the primordial wide binary fraction have to exceed 20% to leave enough wide binaries surviving post - neptune scattering . correcting the estimate of kern & elliot ( 2006 ) to apply to only the cold classical kuiper belt ( using the same approach as lin et al .
2010 ) results in an estimate closer to 5% for the current wide binary fraction , which would imply a primordial fraction of wide binaries in excess of 65% .
these estimates assume no other processes disrupted binaries in the intervening time between implantation in the kuiper belt and the present day .
a primordial wide binary fraction of 20% is comparable to the entire current binary fraction in the cold classical kuiper belt . as mentioned previously , the nesvorn et al .
( 2010 ) simulations appear to over - produce wide binaries with respect to today s population ; synthetic binaries as wide or wider than the ultra - wide tnbs characterized in this work created by the nesvorn et al .
( 2010 ) model represent roughly 20% of the systems produced by their simulations .
therefore , if the primordial binary fraction was @xmath380% , the ultra - wide binary population produced in these simulations would be roughly sufficient to leave a remnant population similar to the observed population after implantation in the cold classical kuiper belt through neptune scattering if no additional processes disrupted these systems afterward .
however , since a much larger fraction of the tighter binaries would survive the implantation process , we would expect a significantly larger remnant population of tight binaries if the primordial binary fraction was so high .
later processes ( such as collisions ) could modify the binary fractions over the age of the solar system , but generally wide binaries are more susceptible to stripping processes and the relative fraction of wide to tight binaries will tend to decrease over time .
therefore , barring a formation mechanism which much more strongly favors wide separations , we find that the ultra - wide binary orbits confirm the findings of parker & kavelaars ( 2010 ) that it is unlikely for the cold classical kuiper belt to have experienced a period of scattering encounters with neptune and likely formed in situ .
in addition to being susceptible to disruption by tidal effects , collisions are effective at unbinding these systems .
petit & mousis ( 2004 ) estimated the mean lifetimes for wide binaries given an estimate of the number of impactors in the current kuiper belt capable of unbinding these systems in an impact .
they found that binaries like 2001 qw@xmath2 had mean lifetimes much less than the age of the solar system , and required a primordial population of wide binaries at least seven times larger than the extant population to explain the continued existence of several such systems . following petit & mousis ( 2004 ) and a.h .
parker & kavelaars ( 2011 , in submitted ) , we estimate the mean lifetime of a binary on a circular mutual orbit due to collisional unbinding by an impact on the secondary to be : @xmath381 where @xmath382 is the number of impactors larger than @xmath383 , drawn from a power - law size distribution with slope @xmath384 , @xmath385 is the mean impact velocity , and @xmath386 is the intrinsic collisional probability . for the following discussion , we adopt the same values as petit & mousis ( 2004 ) for @xmath387 km@xmath388 yr@xmath389 , and @xmath390 km s@xmath389 , which were based upon farinella et al .
( 2000 ) .
this calculation can be somewhat refined by noting that collisions onto the primary can also contribute to unbinding the system , and when folding in this second decay channel the system s mean lifetime becomes half the harmonic mean of the individual collisional lifetimes of the primary and the secondary , @xmath391 for equal - mass systems like 2001 qw@xmath2 , this effectively halves the estimated mean lifetime . for an impactor population ,
we take the estimated total population of tnos in the cfeps l7 model which cross through the cold classical kuiper belt region and extrapolate to small sizes . to do this
, we first estimate the number of objects at @xmath392 ( roughly the magnitude of the putative break in the luminosity function of the kuiper belt ) by extrapolating from the limits of the cfeps survey ( @xmath393 ) with a luminosity function power - law slope of 0.76 ( e.g. , fraser & kavelaars 2009 ) .
we estimate that there are of order @xmath394 objects brighter than @xmath392 in which pass through the cold classical kuiper belt . translating this to a radius with a geometric albedo of 0.1
, we adopt @xmath395 km as our break and normalizing radius for the small - object size distribution . a number of groups ( bernstein et al .
( 2004 ) , fraser & kavelaars ( 2009 ) , fuentes et al .
( 2009 ) ) have found that the luminosity function of the kuiper belt breaks to a shallower slope of approximately @xmath396 , translating to a power - law slope of @xmath397 .
however , fraser ( 2009 ) showed that the size distribution likely steepens again after a `` divot , '' and the true small object size distribution slope may be somewhat steeper than @xmath398 . to capture this uncertainty in the behavior of the size distribution below the break
, we estimate the collisional lifetimes of the ultra - wide tnbs using both a shallow slope of @xmath398 and a collisional - equilibrium slope of @xmath399 .
this steeper slope is comparable to the estimate of the size distribution slope at small sizes set by the putative detection of a single sub - km tno by stellar occultation presented in schlichting et al .
( 2009 ) .
lcccc + + name & & + & @xmath400 ( yr ) & @xmath401 & @xmath400 ( yr ) & @xmath401 + 2000 cf@xmath4 & 5.0@xmath402 & 1 & 1.0@xmath403 & 52 + 2001 qw@xmath2 & 2.6@xmath402 & 1 & 2.7@xmath403 & 4 + 2003 un@xmath34 & 3.3@xmath402 & 1 & 2.5@xmath403 & 5 + 2005 eo@xmath36 & 3.1@xmath402 & 1 & 2.9@xmath403 & 4 + 2006 br@xmath34 & 4.2@xmath402 & 1 & 2.2@xmath403 & 6 + 2006 jz@xmath38 & 3.7@xmath402 & 1 & 3.0@xmath403 & 4 + 2006 ch@xmath0 & 3.9@xmath402 & 1 & 2.7@xmath403 & 5 + @xmath404 : estimate of primordial number of wide binaries with same system properties in order to leave one system surviving today , from @xmath405 with @xmath406 years .
collisional lifetimes from eqn .
[ tau ] for each binary system are presented in table 7 .
this table also contains an estimate of the required initial number of binaries @xmath64 to leave behind one system with a given set of properties after @xmath407 years of collisional bombardment , given @xmath405 . for the shallow small - object size distribution slope @xmath398 , all the binary systems characterized in this work have collisional lifetimes much in excess of the age of the solar system , and
if this slope represents the size distribution then the binary population we see today can be taken as representative of the primordial population . however , in the case of the steeper slope @xmath399 , we see that all the systems have lifetimes less than the age of the solar system , and each binary system represents a member of a decayed initial population ranging from @xmath408 times larger than the current population ( the average required primordial wide binary population for this size distribution is @xmath409 times larger than the extant population ) .
we conclude that if the power - law size distribution slope at small sizes remains shallow , then the extant wide binary population can be taken as representative of the primordial wide binary population .
if this is the case , then the preliminary agreement between the predictions of binary formation by gravitational collapse and the observed orbital distribution of wide tnbs remains valid when considering the primordial wide binary orbital distribution .
the estimates of collisional lifetimes presented here are very simple analytical estimates , and ignore some important effects such as the eccentricity of the binary system , mass loss during impacts , and solar tides .
we have refined these calculations and performed extensive numerical simulations of binary disruption by collisions , and will present the results of this work in an upcoming publication .
the system 2000 cf@xmath4 has by far the shortest collisional lifetime , due to its wide separation and very small component sizes ; in fact , it is the second - widest known tnb ( with respect to @xmath69 ) and its components have the lowest mass of any tno currently measured .
it can be unbound by the smallest impactors of all the binary systems considered here , and is therefore most sensitive to the impactor population at small sizes .
the frequency of systems like 2000 cf@xmath4 in the current kuiper belt is therefore of considerable interest , as they act as powerful probes of the collisional environment .
future surveys will likely identify many such systems , and we discuss the prospects for these surveys in the next section . in the previous sections , we have found that the inclination distribution of the wide binaries indicates a very cold dynamical environment at the time of formation , but formation of the binaries post - accretion through the @xmath359 pathway which should dominate in such conditions can be ruled out due to a lack of preference for retrograde orientations .
we also found that the cold classical kuiper belt was never subjected to a period of close encounters with neptune , and therefore likely formed in situ .
finally , the properties of mutual orbits produced by current simulations of gravitational collapse are consistent with the observed properties of binaries in the cold classical kuiper belt .
together , all these results suggest that formation of the cold classical kuiper belt occurred in situ , with objects forming via rapid collapse of small particles directly into large ( @xmath410 km ) planetesimals through primary accretion driven by turbulent concentration of solids in the protoplanetary disk ( eg . , johansen et al .
2007 , cuzzi et al . 2010 ) .
this process offers a mechanism for forming binaries such as are observed in this population , and can feasibly form the cold classical kuiper belt in situ ( cuzzi et al .
2010 ) without requiring that a large fraction of the primordial mass in planetesimals be subsequently removed through either significant collisional grinding or dynamical processes ( both of which can be destructive to the binary population ) , even when the density of the primordial disk is required to drop sharply beyond 30 au to halt the early migration of neptune at its present distance .
additionally , since this process produces no massive sea of small planetesimals ( most planetesimals being `` born big '' ) , the @xmath359 mechanism will naturally be ineffective , thus preventing it from forming a dominantly retrograde binary population in the dynamically cold environment indicated by the binaries low mutual inclinations
. the mechanism of binary formation by gravitational collapse should be extensively explored . if it can be shown to produce roughly equal numbers of prograde and retrograde binaries while maintaining a preference for low mutual inclinations , then this should be taken as strong support for the formation of the cold classical kuiper belt in situ by rapid primary accretion driven by turbulent concentration and gravitational collapse . while the current sample of known ultra - wide tnbs
remains fairly small , future wide - area surveys like the large synoptic survey telescope ( lsst ) and pan - starrs will be quite capable of identifying such systems .
the projected lsst @xmath411-band single - visit depth is @xmath412 , which translates to @xmath413 for a geocentric distance of 45 au . at this depth
, there are roughly 17,000 objects in the cc - l7 component of the cfeps synthetic model of the kuiper belt . at a minimum , the results of lin et al .
( 2010 ) suggest that at least 1.5% of these objects are ultra - wide tnbs with characteristics like those in our sample , and we therefore expect over 250 such systems will be detectable by the lsst pipeline
. some fraction of the observations of each of these tnbs will likely be resolved , depending on the binary separation and the seeing at the time of each observation . in order to estimate this fraction of resolved observations
, we have used clones of each binary in our sample to determine the distribution of on - sky separations that can be expected over the ten - year baseline of the lsst survey .
we move each clone to a random starting point on both its heliocentric orbit and mutual orbits , then propagated its motion forward for ten years while sampling its on - sky separation at 100 `` observation epochs '' over than ten - year period . at every `` observation epoch '' we determined if the system was resolved or not based on its on - sky separation and an image quality drawn from the measured @xmath6-band distribution of seeing at the lsst site ( lsst science book version 2.0 , section 2.2 ) , with median seeing of @xmath414 .
figure [ lsst ] illustrates the cumulative fraction @xmath415 of binary systems versus the minimum fraction of observations which resolve those systems .
-band taken over the course of the 10-year lsst survey lifetime .
an observation is said to be `` resolved '' when the full - width at half - max of the seeing during that observations is smaller than the projected binary separation.,scaledwidth=50.0% ] the minimum fraction of resolved observations for any system we simulated was @xmath41618% , and the maximum was 100% .
if we take the planned baseline number of 230 visits per field in the @xmath417band to represent the number of observations likely to be made of each binary , then we would expect @xmath41640230 resolved observations of each binary system .
this indicates that virtually _ all _ 250 tnb systems with properties similar to those in our sample likely to be observed by the lsst will have more resolved observation epochs than the most - observed binary in our current data set ( 2001 qw@xmath2 , with 35 observed epochs ) . as such ,
not only will lsst _ detect _ most of these binary systems , but it will measure their mutual orbit properties at least as well as we have managed to do with the sample presented in this paper . with such a large catalog of well - characterized binary orbits , all drawn from the same well - calibrated survey , unprecedented avenues of investigation will be opened .
identifying trends of binary properties with heliocentric orbit may provide further evidence for different origins and histories of the sub - components of the kuiper belt , and will provide new constraints on the dynamical history of the outer solar system ( parker & kavelaars 2010 , murray - clay & schlichting 2011 ) , trends of binary fraction with component sizes will constrain the extent of collisional grinding in the early kuiper belt ( nesvorn et al .
2011 ) , and color albedo trends may be identified and used to constrain the surface composition of these objects .
we have presented the first - ever well - characterized mutual orbits for a sample of seven ultra - wide trans - neptunian binaries .
these orbits range over eccentricities from @xmath418 , and have semi - major axis to hill radius ( @xmath69 ) fractions ranging from @xmath419 .
we find that their properties are distinct from other binary populations , with the following highlights : 1 .
the outer orbits of all widely - separated binaries ( @xmath260 ) with near - equal mass components ( @xmath261 , or @xmath420 ) are distributed consistently with being drawn from the `` stirred '' and `` kernel '' components of the cfeps l7 synthetic model of the kuiper belt , and inconsistent with being drawn from the low - inclination subset of the `` hot '' component .
this confirms that only the dynamically cold components of the classical kuiper belt are host to wide binaries .
two ultra - wide tnbs have very high mutual eccentricities ( @xmath421 0.84 and 0.9 ) consistent with the predictions of binary formation by exchange reactions ( funato et al .
2004 ) , but the rest are inconsistent with the extreme eccentricities predicted by this mechanism .
ultra - wide tnbs have roughly equal numbers of prograde and retrograde orbits , and the observed ratio is inconsistent with the predicted preference for retrograde orbits if the @xmath359 binary formation mechanism dominated unless over @xmath422% of wide binary systems have switched orientations over their lifetimes .
ultra - wide tnbs have a statistically significantly different mutual inclination distribution compared to tighter binaries in literature , and their inclinations also can not have been drawn from a uniform sphere ( @xmath291 ) due to a lack of detections of high - mutual inclination wide binaries .
this paucity of wide binaries at high inclinations and preference for low inclinations can not be explained by the kctf mechanism alone , and suggests a primordial preference for low - inclination mutual orbits .
this suggests formation in a very dynamically cold disk .
the wide separations of these systems indicates that if binary formation was dominated by the @xmath360 mechanism , the velocity dispersion in the primordial disk must have been less than a few times the hill velocity .
a low velocity dispersion supports the observed preference for low mutual inclinations .
however , the velocity dispersion could not have been much below the hill velocity or the @xmath359 mechanism would have dominated and the ultra - wide binaries should exhibit a strong preference for retrograde orbits .
thus , under the assumption that these are the only two efficient formation pathways , it appears that @xmath369 .
further modeling of these mechanisms is required to determine whether a balance can be struck between them where widely - separated binaries can be formed with aligned poles and roughly equal numbers of prograde and retrograde orientations .
current simulations of the alternative formation mechanism of gravitational collapse and fragmentation ( nesvorn et al .
2010 ) create orbital distributions similar to the observed present - day distributions .
these simulations tend to over - produce wide systems , which leaves some room for post - formation disruption of ultra - wide binaries .
however , it remains unclear if this mechanism can produce a large number of retrograde systems .
7 . assuming realistic densities , the implied albedos for the ultra - wide tnbs range over @xmath423 , consistent with estimates of the albedos of solitary cold classical objects ( eg .
, brucker et al .
the distribution of observed albedos with primary radius for all dynamically - cold tnbs ( seven from this work and four others from literature ) suggests that albedos at the lower end of this scale are intrinsically more common in this population . a gaussian albedo distribution , centered at @xmath352 and clipped such that @xmath424 , is consistent with observations for widths @xmath425 .
this estimate does not account for the possibility that binary fraction might vary substantially with primary radius as predicted by nesvorn et al .
all seven systems characterized in this work are widely separated enough to have a substantial probability of disruption if ever subjected to close encounters with neptune ( parker & kavelaars 2010 ) , suggestive of _ in situ _ formation or a more gentle migration mechanism .
additionally , the collisional lifetimes of these binary systems are short if the size distribution of the kuiper belt is steep ( @xmath426 ) at small sizes ; for the present population of binaries to have survived for the age of the solar system requires that they were never subjected to a period of intense collisional grinding ( eg . , petit & mousis 2004 , nesvorn et al .
2011 ) and that significant collisions have remained relatively rare in the kuiper belt over the age of the solar system .
we explore the collisional evolution of these binary systems and further constrain the collisional environment of the present - day kuiper belt in an upcoming paper . 9 .
a confluence of results _ ( a ) _ gravitational collapse producing similar binary properties as are observed , _ ( b ) _ the evidence that the cold classical kuiper belt was not subjected to a period of close encounters with neptune , and _
( c ) _ the mutual orbits of the wide binaries suggesting formation in a very dynamically - cold environment , yet the @xmath359 mechanism not dominating the binary formation process all suggest that the cold classical kuiper belt formed in situ through rapid accretion of small particles directly into large ( @xmath410 km ) planetesimals , driven by turbulent concentration of solids and gravitational collapse .
we thank melissa graham for pointing out that the malmquist bias lurks everywhere , and david nesvorn for providing the results from his simulations of binary formation .
we also thank the gemini staff for their technical support of our observing programs .
alex parker is funded by the nsf - grfp award dge-0836694 .
this research used the facilities of the canadian astronomy data centre operated by the national research council of canada with the support of the canadian space agency .
gladman , b. , marsden , b. g. , & vanlaerhoven , c. 2008 , in the solar system beyond neptune , ed .
m. barucci , h. boehnhardt , d. p. cruikshank , & a. morbidelli , ( tucson , az : univ .
arizona press ) , 592 , 43 - 57 .
noll , k. , grundy , w. , chiang , e. , margot , j .-
& kern , s. 2008a , in the solar system beyond neptune , ed .
m. barucci , h. boehnhardt , d. p. cruikshank , & a. morbidelli , ( tucson , az : univ .
arizona press ) , 592 , 345 - 363 .
petit , j .-
m . , kavelaars , jj .
, gladman , b. , margot , j .-
nicholson , p. , jones , r. , parker , j. , ashby , m. , campo - bagatin , a. , benavidez , p. , coffey , j. , rousselot , p. , mousis , o. & taylor , p. 2008 .
science 322 , 432 . | the low - inclination component of the classical kuiper belt is host to a population of extremely widely - separated binaries .
these systems are similar to other trans - neptunian binaries ( tnbs ) in that the primary and secondary components of each system are of roughly equal size .
we have performed an astrometric monitoring campaign of a sample of seven wide - separation , long - period tnbs and present the first - ever well - characterized mutual orbits for each system .
the sample contains the most eccentric ( 2006 ch@xmath0 , @xmath1 ) and the most widely - separated , weakly bound ( 2001 qw@xmath2 , @xmath3 ) binary minor planets known , and also contains the system with lowest - measured mass of any tnb ( 2000 cf@xmath4 , @xmath5 kg ) .
four systems orbit in a prograde sense , and three in a retrograde sense .
they have a different mutual inclination distribution compared to all other tnbs , preferring low mutual - inclination orbits .
these systems have geometric @xmath6-band albedos in the range of @xmath7 , consistent with radiometric albedo estimates for larger solitary low - inclination classical kuiper belt objects , and we limit the plausible distribution of albedos in this region of the kuiper belt . we find that gravitational collapse binary formation models produce a similar orbital distribution to that currently observed , which along with a confluence of other factors supports formation of the cold classical kuiper belt _ in situ _ through relatively rapid gravitational collapse rather than slow hierarchical accretion . we show that these binary systems are sensitive to disruption via collisions , and their existence suggests that the size distribution of tnos at small sizes remains relatively shallow . |
the random trap model together with its close relative , the barrier model , can be applied to a variety of physical problems @xcite which are related to random walks in disordered media , like properties of a photocurrent in amorphous solids @xcite , or have an equivalent mathematical representation , like the behavior of resistor and capacitance networks @xcite or lattice vibrations of harmonic chains @xcite .
it has also become prominent as a toy model for the phase space dynamics of spin glasses @xcite .
the random trap model on a regular lattice consists of a trapping landscape @xmath0 of energy wells located on the lattice s sites @xmath1 .
the @xmath2 are iid random variables which in our case have the common density @xmath3 with @xmath4 being the characteristic temperature . according to the vant hoff - arrhenius law @xcite , at a temperature @xmath5 these trapping potentials
correspond to _ mean waiting times _ @xmath6 , where we put @xmath7 in what follows .
the density of the mean waiting times @xmath8 is @xmath9 with @xmath10 . in
what follows , we consider lattices in dimensions @xmath11 and @xmath12 . in 1d
the particle s dynamics on the lattice is described by a continuous - time markov chain with transition rates @xmath13 with @xmath14 being the number of neighbors of site @xmath1 and @xmath8 denoting the mean waiting time associated with the site .
if the first moment of the distribution , eq.([equ : rhotau ] ) , diverges , which is the case for @xmath15 , the dynamics of the system starts to show highly nontrivial behavior ( as exemplified by anomalous diffusion in such models ) .
of course , one could also define a mixture of the trap and barrier model , or generalize the model as in @xcite .
the model above describes the behavior of random walkers who , while exploring the lattice , get trapped in the potential wells .
once trapped at site @xmath1 , they have to wait for a time @xmath16 ( actual waiting time ) with pdf @xmath17 containing the mean waiting time @xmath8 at a site @xmath1 as parameter .
note that these local means @xmath8 are distributed according to eq .
( [ equ : rhotau ] ) and the exponential waiting time density of eq .
( [ equ : exponentialpdf ] ) is already conditioned on the fact that a random walker resides on a site with mean waiting time @xmath8 .
after the random walkers have waited long enough , they jump with equal probability to any of the recent site s neighbors .
thus , the spatial properties of each single random walker s trajectory are the same as in a simple random walk on the lattice . what is interesting , is the temporal evolution .
the single - particle ( one - time ) quantities such as the mean squared displacement are well - understood , and can in high dimensions be described by the continuous - time random walk scheme ; the behavior in 1d , which shares many properties with the ctrw behavior , is also well - investigated @xcite .
the properties of the motion of two or more random walkers exploring the same energy landscape are much less understood . in what follows we concentrate on some of these multiparticle quantities , from which the ( inverse ) participation ratio @xmath18 , which represents the probability that two random walkers who started at @xmath19 at the same site @xmath20 meet again at time @xmath21 , was considered the most interesting .
if the lattice is finite , the disorder - averaged long - time limit @xmath22 converges to an equilibrium value @xmath23 @xcite , independent of the lattice s dimension @xmath24 .
the question about the non - equilibrium behavior of @xmath25 , however , is much more involved .
it seems especially fascinating that in dimension one even for infinitely large lattices there exist a finite disorder - averaged long - time limit @xmath26 with @xmath27 , the phenomenon called _
dynamical localization_. this manifests itself also in finite but large one - dimensional lattices as a pronounced plateau in the temporal evolution of @xmath28 ( see fig . [
fig : y2p0av ] ) .
the existence of this regime has been mathematically proven by fontes _
_ @xcite and was further investigated by bertin and bouchaud @xcite . for lattices with dimension @xmath29 , fontes _
@xcite proved that the dynamical localization is absent ; it is conjectured that it is also absent in @xmath30 .
the phenomenon of dynamical localization is thus a property of one - dimensional lattices , i.e. : @xmath31 the order of the limits is crucial since otherwise the expression converges to the equilibrium value @xmath32 .
the goal of this paper is to study the dynamical localization in somewhat more detail , and to connect it with the properties of the spectrum and of the eigenfunctions of the laplacians defined by eq .
( [ equ : generalrates ] ) in the one- and three - dimensional lattices with periodic boundary conditions .
this is done by extensive numerical simulations relying on the spectral representation of the dynamics .
the method allows for a numerically exact calculation of the system s properties averaged over the initial position of the walkers and their thermal histories ( trajectories ) for a given realization of traps , and for considering longer observation times than the direct monte - carlo simulations .
apart from investigating the behavior of @xmath33 , we consider some other measures of localization , as well as the properties of the eigenfunctions of the corresponding laplacians , trying to hint onto the main properties of the system responsible for dynamical localization .
( dashed lines ) and @xmath34 ( dots ) for @xmath35 , @xmath36 and the disorder average over 100 realizations .
the light gray area gives the whole range of @xmath37 of all 100 realizations and the dark gray area depicts @xmath38 with @xmath39 being the standard deviation . ]
let us consider a 1d ring of @xmath40 sites and a realization @xmath41 of waiting times according to the prescribed waiting time distribution @xmath42 of eq .
( [ equ : rhotau ] ) .
let @xmath43^n$ ] be the vector whose components @xmath44 describe the probability to find the random walker at site @xmath1 at time @xmath21 . then the participation ratio @xmath33 is simply @xmath45 .
the temporal evolution of @xmath46 is given by the matrix @xmath47 of the transition rates @xmath48 which are explicitly defined by eq .
( [ equ : generalrates ] ) .
all eigenvalues of @xmath47 are nonpositive .
in what follows we will refer to the laplacian matrix ( laplacian ) @xmath49 , with @xmath50 and nonnegative eigenvalues .
the master equation thus reads @xmath51 and it leads to the temporal evolution of @xmath46 in the form @xmath52 with @xmath53 being the initial condition . in our case , the initial distribution @xmath53 will always be completely localized , i.e. @xmath54 .
if we average over all these localized initial conditions , then @xmath55 becomes the trace @xmath56 with @xmath57 being the @xmath58-th eigenvalue of the time - dependent operator @xmath59 . in what follows we will however use a more convenient representation and describe the process in the basis of the eigenfunctions of the laplacian @xmath60 where the spectral decomposition is time - independent .
if @xmath61 is the laplacian for a simple random walk on a ring of size @xmath40 ( i.e. a toeplitz matrix with @xmath62 on the main diagonal , minus one half on the two adjacent diagonals and in the upper right and lower left corners , and all other elements equal to zero ) and @xmath63 , the laplacian @xmath60 is given by @xmath64 the transpose of @xmath60 is @xmath65 and the matrix @xmath66 is symmetric and hence diagonalizable over @xmath67 .
this diagonalizability is inherited to @xmath60 and @xmath68 .
we will denote the eigenvectors of @xmath60 by @xmath69 , of @xmath70 by @xmath71 and of @xmath68 by @xmath72 . since @xmath70 is symmetric
, we can choose the @xmath71 such that they form an orthonormal basis of @xmath73 and define the eigenvectors of @xmath60 and @xmath68 such that for eigenvectors to the same eigenvalue @xmath74 the following relation holds : @xmath75 from this follows : @xmath76 where we have used the fact that @xmath77 is diagonal .
furthermore , there is an equilibrium state of @xmath60 , namely the eigenvector @xmath78 to eigenvalue zero , corresponding to the eigenvector @xmath79 of @xmath68 .
all other eigenvalues are strictly positive since @xmath80 with @xmath77 being invertible and @xmath61 being ( symmetric ) positive - semidefinite with exactly one vanishing eigenvalue . the eigenvalues can be ordered such that @xmath81
we want to express the temporal evolution of @xmath82 in terms of eigenvectors and eigenvalues of @xmath60 and @xmath68 .
it is easily verified that @xmath83 and thus @xmath84 if we average over all initial conditions @xmath53 fulfilling @xmath54 , then @xmath85 with @xmath86 the @xmath87 identity matrix . note , that we have not yet applied any disorder average . averaging eq .
( [ equ : ptpinitial ] ) over the initial conditions and applying eq .
( [ equ : p0average ] ) we obtain @xmath88 where the elements @xmath89[2em][c]{0 cm } { } } \\ & = ( { \mathbf{q } } _
i^t s^2 { \mathbf{q } } _ j ) ( { \mathbf{q } } _
i^t s^{-2 } { \mathbf{q } } _ j ) { \parbox[0pt][2em][c]{0 cm } { } } \nonumber , \end{aligned}\ ] ] define a new matrix @xmath90 which is the element - wise product of the gramian matrices corresponding to the vector sets @xmath91 and @xmath92 . thus , the temporal development of @xmath37 depends on @xmath93 in relation to the characteristic decay times @xmath94 .
as we proceed to show , these are the properties of @xmath90 which dominate the dynamical localization . taking the temporal limit of eq .
( [ equ : ptpaveraged ] ) in a finite system and averaging over disorder , we can also reproduce the finite - size equilibrium value @xmath95 as it was given in @xcite : @xmath96 see @xcite .
( [ equ : ptpaveraged ] ) comes in very handy because it enables us to determine @xmath37 at any time @xmath21 without previously computing everything that happened before @xmath21 .
additionally , the temporal behavior of @xmath37 ( which is still dependent on the disorder ) can be simulated quite efficiently , i.e. without the statistical error attached to the average over initial conditions and over single trajectories .
thus the matrix approach gives a useful numerical tool of investigating the problem of dynamical localization , the tool which will be continuously used in the present paper .
a nice consequence is that we are able to examine the temporal behavior of @xmath37 as averaged over initial conditions and trajectories in a fixed disorder realization and compare it to the disorder - averaged behavior to evaluate the inter - sample differences , as is done in fig .
[ fig : y2p0av ] . fig .
[ fig : y2p0av ] shows the behavior of @xmath37 in three different realizations of the random potential together with the result @xmath97 as following from averaging over 100 such realizations .
the pronounced plateau of @xmath97 between @xmath98 and @xmath99 corresponds to the dynamical localization and up to now was only considered after applying the disorder average .
we see that up to the crossover at @xmath100 the @xmath101 do not differ strongly from each other , so that the hints onto the dynamical localization are present in each single realization of the random traps . especially at the onset of the plateau at @xmath102
all different @xmath101 appear to follow the same pattern , independent on what equilibrium value of @xmath28 would be attained .
[ fig : y2p0av ] shows that , even though the temporal evolution of the sample @xmath37 fluctuates around the disorder average @xmath34 , the effect of the dynamical localization is not an exclusive property of @xmath34 but does already appear in the average over initial conditions and thermal histories .
the analytical calculation of @xmath33 in the non - equilibrium case pertinent to dynamical localization turned out to be quite difficult .
bertin and bouchaud @xcite conjectured @xmath103 using a simple approximation and presented simulations with infinite - time extrapolation that confirmed their conjecture .
we find it worth mentioning that according to their simulations ( * ? ? ?
5 ) as well as in our attempts to reproduce their figure , the simulated data tend to be smaller than eq .
( [ equ : bbconjecture ] ) predicts , whereas it is more likely that the fitting procedures overestimate the infinite - time value @xmath104 .
nevertheless , eq . ( [ equ : bbconjecture ] ) gives a simple and elegant approximation .
the localization of states was extensively studied for vibrational excitations and in quantum models , where different localization criteria were applied mostly to the eigenstates of the corresponding hamiltonians ; in a typical case all of them lead to the same kind of conclusions on whether the corresponding state is localized or not .
thus , the participation ratio quantifies to what extent the localized state is _ concentrated _ on a small subset of all sites . the standard criterion for localization which is both theoretically transparent and easy for numerical implementation , especially in one dimension ,
is based on the evaluation of the lyapunov exponent @xmath105 which corresponds to the rate of exponential growth of the amplitude with the coordinate @xcite , and determines the behavior of the state outside of its localization region .
we will consider the behavior of the lyapunov exponent later on in this paper .
one can also quantify the localization by the radius of gyration @xmath106 of the corresponding state , which indicates the spatial _
spread _ of the state , the size of the spatial domain which the particle or excitation is confined to . of the radius of gyration in dependence of the system size . in agreement with eq .
( [ equ : rg2eq ] ) , the simulation results ( dots ) follow @xmath107 ( dashed line ) . ] it is defined as @xmath108 with @xmath109 denoting the distance between sites @xmath1 and @xmath110 . thus , the gyration radius gives the mean squared distance between the two diffusion particles . in a linear lattice with closed boundary conditions @xmath111 . in the case of a ring of size @xmath40 the distance is given by @xmath112 we say a state is localized if the radius of gyration is small . with the definition of the matrix @xmath113 with the components @xmath114 we find : @xmath115 ( ) in comparison to @xmath116 ( * + * ) , calculated as average over 100 landscape realizations of size @xmath36 and @xmath35 . ]
this has the same structure as @xmath117 with @xmath118 being the identity matrix .
but , whereas the bilinear form @xmath118 is a diagonal matrix , @xmath119 has vanishing diagonal entries and is dominated by non - diagonal ones . in analogy to the calculations that led to eq .
( [ equ : ptpaveraged ] ) , we find : @xmath120 the equilibrium value in disorder average can be computed easily : @xmath121 the two factors can now be evaluated independently : @xmath122 where we have used the results of @xcite .
the evaluation of @xmath123 yields @xmath124 so that @xmath125 fig .
[ fig : rg2ofn ] shows the results of simulations of eq .
( [ equ : rg2averaged ] ) in the limit @xmath126 for different @xmath127 and system sizes @xmath40 and the comparison to eq .
( [ equ : rg2eq ] ) .
the important information is that the equilibrium value of the radius of gyration grows with the system size whereas the equilibrium value of @xmath25 becomes independent of @xmath40 ( for @xmath40 large enough ) , i.e. the states localized according to the @xmath33 criterion are delocalized with respect to their gyration radius .
the reason is quite clear .
if a site attracts a considerable amount of probability in the equilibrium state @xmath128 , let us call it a `` deep trap '' . if there is only one deep trap in the landscape
, @xmath33 is large , and @xmath106 is small .
many landscape realizations do contain _ more than one _ deep trap , in which case @xmath33 is still large , but @xmath106 might be large as well , being of the size of the distance between the deep traps .
the expectation of the distance between the deep traps grows with the system size and thus the radius of gyration does also grow , whereas the participation ratio is independent of the deep traps positions .
apart from the equilibrium value , we can also compare the temporal evolution of @xmath116 to the one of @xmath34 .
[ fig : rg2temporal ] depicts such a comparison with numerical averages over 100 landscape realizations .
altogether , the phenomenon of dynamical localization is pertinent to the behavior of one , very specific localization measure . in the range of times
when the dynamical localization is observed , the probability that the two independent random walkers starting at the same site appear to be at the same site at a later instant of time @xmath21 ( as represented by @xmath33 ) stagnates , but the mean squared distance between them ( given by @xmath106 ) grows continuously .
@xmath33 is thus a measure of the concentration of the state on some small subset of sites ; @xmath106 , on the other hand , quantifies the size of the spatial domain in which the particles are confined .
both the dynamically localized state and the equilibrium state of the system are essentially delocalized with respect to this measure . vs. inverse radius of gyration @xmath129 ( see eq .
( [ equ : sigma-2 ] ) ) on double logarithmic scales .
the shaded area gives the density of single data points , whereas the dots represent averaged values .
the larger circles define the ensemble for a linear fit ( solid line ) , yielding @xmath130 .
the dashed line represents the numerical zero @xmath131 . ]
the equilibrium state is essentially an eigenvector of @xmath60 to its zero eigenvalue ; therefore it is interesting to look at the eigenvalue localization properties of @xmath60 also for other eigenvalues @xmath132 .
the corresponding measure of localization is given by the lyapunov exponent @xmath105 which is defined as follows : let @xmath44 be the components of an eigenvector of @xmath60 to eigenvalue @xmath132 .
then the following relation holds : @xmath133[2em][c]{0 cm } { } } \\ \frac{\pi_{k+1}}{\tau_{k+1 } } { \parbox[0pt][2em][c]{0 cm } { } } \end{pmatrix } & = \underbrace{\begin{pmatrix } 0 & 1\\ -1 & 2 - \lambda \tau_k \end{pmatrix}}_{=~t_k } \begin{pmatrix } \frac{\pi_{k-1}}{\tau_{k-1 } } { \parbox[0pt][2em][c]{0 cm } { } } \\ \frac{\pi_k}{\tau_k } { \parbox[0pt][2em][c]{0 cm } { } } \end{pmatrix}. \label{equ : tk2}\end{aligned}\ ] ] the matrix @xmath134 which transfers between the vector on the r.h.s . and the vector on the l.h.s .
is called the @xmath1th transfer matrix .
the lyapunov exponent is now defined as @xmath135 with @xmath136 denoting the maximum absolute column sum of the matrix @xmath137 .
( [ equ : defgamma ] ) together with eq .
( [ equ : tk2 ] ) defines @xmath138 regardless of the fact whether @xmath132 is an eigenvalue of @xmath60 or not .
( note that the vectors in eq .
( [ equ : tk2 ] ) are actually built from the components of the eigenvectors of @xmath68 . in the limit @xmath139
, however , it can be shown that the lyapunov exponent @xmath138 of both @xmath60 and @xmath68 is the same . )
as the lyapunov exponent gives the rate of exponential growth in an ( eigen-)vector , it is a standard measure for localization : @xmath140 means that the state is delocalized whereas a high lyapunov exponent means strong localization . considering this
, it seems peculiar to us that the equilibrium state of the trap model should have a high value of participation ratio but a vanishing lyapunov exponent @xmath105 : , @xmath35 , the number of realizations is 100 . ] @xmath141 thus , the equilibrium state is essentially delocalized with respect to the its lyapunov exponent .
since in all our simulations we consider finite systems , the measured lyapunov exponent at @xmath142 is of order @xmath131 . a double - logarithmic plot of the lyapunov exponent for @xmath143 and @xmath144 is shown in fig .
[ fig : gammaofecloseup ] . for all four values of @xmath145 the lyapunov exponent @xmath146 is of order @xmath131 ( dashed line ) which corresponds to zero in the infinite system size limit
. the lyapunov criterion of localization gives information which is strongly correlated with the one delivered by the gyration radius of single eigenstates .
the inverse squared gyration radius for an eigenstate can be defined via @xmath147 and @xmath148 being the @xmath58th entry of the @xmath1th eigenvector . in fig .
[ fig : gammaofsigma ] we plot @xmath149 vs. @xmath150 on double logarithmic scales .
strong correlation ( linear on double logarithmic scales , i.e. a power - law dependence ) between both is seen .
a linear fit shows that @xmath130 .
this means that the lyapunov exponent and the gyration radius provide related information on the eigenstate localization : the eigenstates confined to a smaller region typically decay faster outside of it , and vice versa .
now we turn to another important question , namely where exactly the information on the dynamical localization is hidden : is it determined by the spectral properties of the laplacian matrix , or is it mostly coded in the entries of the matrix @xmath90 composed from the eigenvectors of @xmath60 ? to understand this we compare the behavior of a one - dimensional system of @xmath40 sites with the one of its three - dimensional analogue , where dynamical localization is absent @xcite .
the total number of sites shall also be @xmath40 , i.e. the side length of the cube is @xmath151 .
the rates for the 3d case are as follows : @xmath152 a comparison between the disorder averaged evolution of @xmath33 on the one- and three - dimensional case is shown in fig .
[ fig : y21d3d ] . as fontes
@xcite proved , there is no dynamical localization in the 3d case , i.e. @xmath153 , which , for finite lattices with site number @xmath40 , can be understood as @xmath154 . in fig .
[ fig : y21d3d ] this can be seen by the fact that @xmath155 is of the order of @xmath156 and @xmath157 is considerably larger although in both cases the number of sites is @xmath36 . in fig .
[ fig : spectrum ] we plotted the spectrum @xmath158 for both laplacians @xmath60 and @xmath159 on linear and on double - logarithmic scales .
note that the cusp at @xmath160 is due to the fact that we use pure pareto - distributed waiting times and it does not show up for analytic distributions as for example a one - sided lvy - distribution with corresponding asymptotics . in any case , the behavior of @xmath158 for large values of @xmath132 does not seem interesting to us because we are looking for hints for the long - time behavior of the system which means we are concerned about small @xmath132 . the double - logarithmic plot of fig .
[ fig : spectrum ] reveals that for small values of @xmath132 , the spectrum @xmath158 exhibits a power - law behavior .
the exponent in the 3d case is in fact equal to @xmath161 , as the dashed line in the double - logarithmic plot indicates .
this is the exponent bovier and faggionato @xcite determined for infinite - dimensional models .
note , that the dashed line in fig .
[ fig : spectrum ] is not a fit .
the absolute value of the exponent for the 1d case is slightly larger than in 3d . ) and ( [ equ:1dwith3deigenvalues ] ) for @xmath36 , @xmath35 and 100 landscape realizations .
the exact result for the 1d case is given by the solid line and the approximation of eq .
( [ equ : factorized ] ) is depicted by circles , the 3d case by plus signs .
the approximation of eq .
( [ equ:1dwith3deigenvalues ] ) is represented by squares . ] nevertheless , the qualitative behavior of both the spectrum of @xmath60 and @xmath159 does not differ strongly .
the question is , whether the different exponents of @xmath158 do contribute to the effect of dynamical localization , or whether they are only an expression of the different time scales of the corresponding random walks . in order to investigate this , as well as the interplay between the eigenvectors and eigenvalues , we test the following decoupling approximation of eq .
( [ equ : ptpaveraged ] ) : @xmath162 with @xmath163 .
the result of this approximation is depicted in fig .
[ fig : factorized ] .
the figure shows that the decoupling approximation adequately reproduces the results of direct simulations and that the correlations between eigenvalues and the entries of @xmath90 do not actually play a role .
entries of the matrix @xmath90 defined by eq .
( [ equ : defg ] ) .
@xmath143 , @xmath35 , 100 realizations .
gray bars represent positive entries and white bars negative ones.,title="fig : " ] entries of the matrix @xmath90 defined by eq .
( [ equ : defg ] ) .
@xmath143 , @xmath35 , 100 realizations .
gray bars represent positive entries and white bars negative ones.,title="fig : " ] to elucidate the the influence of the different spectra in the 1d and 3d case , we simply substitute the eigenvalues of the 1d case by the ones of the 3d case ( i.e. use 1d @xmath90-matrix together with the corresponding 3d eigenvalues ) : @xmath164 thus , in this approximation , the only difference between the 1d and the 3d case is the shape of the matrix @xmath90 defined by eq .
( [ equ : defg ] ) .
the result of the approximation eq .
( [ equ:1dwith3deigenvalues ] ) is also presented in fig .
[ fig : factorized ] .
although the plateau of the case eq .
( [ equ:1dwith3deigenvalues ] ) lies lower than the one of eq .
( [ equ : factorized ] ) , it is still there and pronounced .
thus , the exact form of the spectrum is only responsible for the height and the duration of the plateau , not for the fact of its presence or absence .
we thus conclude that the difference in the spectrum @xmath158 is primarily responsible for the different time scales but _ not _ for the fact of dynamical localization . since we have shown that it is the matrix @xmath90 built from the eigenvectors of @xmath60 which codes for the presence or absence of dynamical localization , the question arises , what properties of its elements are mainly responsible for it .
the elements of @xmath90 differ strongly in their magnitude .
is the plateau visible in fig .
[ fig : factorized ] the result of a complex interplay between positive and negative small components of @xmath90 or is its presence dominated by the behavior of large components of @xmath90 other than @xmath165 ( coding for the equilibrium ) ? to attack this question , the first @xmath166 entries of @xmath167 in the 1d and 3d case are depicted in fig .
[ fig : gaveraged ] . in order to make the important information visible
, the @xmath168-axis represents the absolute values of the elements , @xmath169 , whereas the sign of the entries is indicated by the color of the bars : gray bars represent positive entries and white bars negative ones .
since the equilibrium state is the same in 1d and 3d , @xmath170 in both cases . in the 3d case
however , the values of the other diagonal entries decay more rapidly than in the 1d case .
moreover , while in the 3d case , for @xmath171 and @xmath172 , the entries @xmath173 are close to zero and of fluctuating sign , in 1d case all shown off - diagonal entries are negative ( although for larger indices @xmath174 they also fluctuate ) .
( filled symbols ) and @xmath175 ( empty symbols ) of eq .
( [ equ : splittedsum ] ) for the 1d ( circles ) and 3d ( squares ) case .
simulations with system size @xmath143 , @xmath35 , and 100 landscape realizations . ]
let us split the diagonal and the non - diagonal contributions to the overall sum over eigenstates : @xmath176 the diagonal elements of @xmath90 are always positive , and so is @xmath70 . according to the simulations , the term @xmath177 is negative such that @xmath175 is positive .
[ fig : factorization ] shows the temporal evolution of @xmath70 and @xmath175 in both 1d and 3d case , where in both cases we have used again the spectrum of the 3d system .
it seems that the plateau in fig .
[ fig : factorized ] is indeed the result of an interplay of the positive diagonal entries in @xmath70 and the mostly negative ones in @xmath177 . in 1d
both terms decay logarithmically but the sum @xmath178 stays constant over a long time . in three dimensions
the terms behave completely different .
this shows that the presence or absence of dynamical localization is fully coded in the matrix @xmath90 , i.e. depends on the properties of eigenvectors of @xmath60 , and not so much on its spectrum .
it is strongly dominated by the scalar products of eigenvectors to different eigenvalues ( which are not orthogonal since @xmath60 is not symmetric ) entering the non - diagonal terms , and in this sense indeed has to do with eigenvector localization , although the direct relation is not clear yet .
in the present work we applied an algebraic approach to investigate the phenomenon of dynamical localization in the random trap model with power - law distributed mean waiting times . apart from its formal elegance , the approach works extremely well as a computational scheme , since it allows for obtaining numerically exact results for the participation ratio @xmath33 averaged over initial conditions and thermal histories in a given system s realization . only a single averaging procedure over the samples
is necessary .
because this approach enables us to compute the relevant properties at any given time , it allows for considering longer observation times than in the direct monte - carlo simulations .
the phenomenon of dynamical localization , as observed in the participation ratio , is a very peculiar property of one - dimensional trap models .
its physical interpretation is that the probability , that two particles with the same starting position are found together , is surprisingly high for intermediate time scales . in this context , _ intermediate time scales _ means that full equilibration is not yet established .
the fact , that the particles often meet at the same site , does not mean that they stay close to each other : the typical distance between the particles continuously grows with time until the terminal equilibration is reached .
moreover , the equilibrium state , which is the strongest localized one with respect to @xmath33 , is essentially delocalized with respect to its radius of gyration and its lyapunov exponent .
numerical simulations show that the later two are well - correlating localization measures for the eigenstates of the system .
we have moreover shown , that the phenomenon of dynamical localization is only marginally connected with the spectral properties of the laplacian operator governing the system s dynamics , and is dominated by the properties of the eigenfunctions of @xmath60 , which outlines the direction of further investigations .
the financial support of dfg within the irtg 1740 `` dynamical phenomena in complex networks : fundamentals and applications '' is gratefully acknowledged . | the one - dimensional random trap model with a power - law distribution of mean sojourn times exhibits a phenomenon of dynamical localization in the case where diffusion is anomalous : the probability to find two independent walkers at the same site , as given by the participation ratio , stays constant and high in a broad domain of intermediate times .
this phenomenon is absent in dimensions two and higher . in finite lattices of all dimensions
the participation ratio finally equilibrates to a different final value .
we numerically investigate two - particle properties in a random trap model in one and in three dimensions , using a method based on spectral decomposition of the transition rate matrix .
the method delivers a very effective computational scheme producing numerically exact results for the averages over thermal histories and initial conditions in a given landscape realization . only a single averaging procedure over disorder realizations
is necessary .
the behavior of the participation ratio is compared to other measures of localization , as for example to the states gyration radius , according to which the dynamically localized states are extended .
this means that although the particles are found at the same site with a high probability , the typical distance between them grows .
moreover the final equilibrium state is extended both with respect to its gyration radius and to its lyapunov exponent .
in addition , we show that the phenomenon of dynamical localization is only marginally connected with the spectrum of the transition rate matrix , and is dominated by the properties of its eigenfunctions which differ significantly in dimensions one and three . |
non - gaussian fluctuations in mesoscopic devices and in photon statistics have proved to be a powerful probe of the underlying physics of such systems , yielding information that can not be deduced from the averaged value and the variance.@xcite experimentally , detecting a deviation from the gaussian distribution is a challenging task .
direct measurements of the third and higher moments of fluctuations in mesoscopic systems require highly sophisticated techniques.@xcite in another approach , following theoretical proposals,@xcite non - gaussian statistics was detected using noise - activated switching out of a metastable state in a josephson junction.@xcite this approach provides direct access to the tail of the noise distribution , as a relatively large noise outburst is required to drive the josephson junction out of the effective potential well . in the measurements ,
deviations from the gaussian statistics were found from the asymmetry of the switching rate with respect to the current polarity .
noise - induced switching has attracted much attention recently in the context of nonequilibrium systems , where the notion of escape from a potential well does not apply . in particular , a number of experiments were done on periodically driven systems , with the metastable states corresponding to forced vibrations .
the systems ranged from trapped electrons@xcite to josephson junctions,@xcite mechanical resonators , @xcite and trapped atoms.@xcite it was found that , for gaussian noise , the switching rate @xmath0 displays universal features .
its dependence on the noise intensity @xmath1 is of the activation form , @xmath2 , with the switching exponent @xmath3 , where @xmath4 is the effective activation barrier .
furthermore , @xmath4 displays the predicted @xcite scaling behavior close to bifurcation points where the number of stable states changes,@xcite @xmath5 , where @xmath6 is the distance to the bifurcation point . in this paper
we use a driven micromechanical oscillator to study the features of poisson - noise induced switching .
in contrast to gaussian noise , poisson noise is characterized not just by its intensity @xmath7 and correlation time , but also by the pulse rate @xmath8 .
we limit ourselves to the most interesting case of short pulses and explore the dependence of the switching rate on the pulse area and @xmath8 as well as the distance to the bifurcation point @xmath6 .
even though the stable states of this nonequilibrium system are not separated by a potential barrier , our findings are consistent with the recently predicted universal behavior of poisson - noise induced switching near a bifurcation point , which differs qualitatively from that for gaussian noise @xcite . (
a ) schematic of the experimental setup .
the noise force is generated by applying voltage @xmath9 to the top sidegate , with @xmath10 .
( b ) vibration amplitude vs. driving frequency ; the vibration amplitude is scaled by its value at the critical point where hysteresis emerges .
the curves correspond to the driving amplitudes of @xmath11 , @xmath12 , and @xmath13 , where @xmath12 is the critical driving amplitude for the onset of hysteresis .
the up - arrow indicates the position of the bifurcation point @xmath14 .
we fit the linear and the critical responses to extract the parameters of the oscillator .
( c ) schematic of the generation of poisson pulses .
each output pulse contains @xmath15 400 rf cycles . for clarity , only 2 rf cycles are shown.,width=245 ]
our oscillator is a double - clamped polysilicon beam that is @xmath16 m long and @xmath17 m by @xmath18 m in cross section . by passing an ac current through the beam in a perpendicular magnetic field of 5 t
, we excited motion of the beam in its in - plane fundamental mode [ fig .
this motion generates an electromotive force that changes the transmitted ac power .
all measurements were performed at @xmath19 k and @xmath20 torr .
the mode can be modeled by a duffing oscillator @xcite , @xmath21 where @xmath22 is the normalized beam displacement , @xmath23 rad@xmath24s@xmath25 is the damping coefficient , @xmath26 rad@xmath24s@xmath25 is the resonant frequency , @xmath27 is the coefficient of the cubic nonlinearity , @xmath28 and @xmath29 are , respectively , the amplitude and frequency of the external driving force , and @xmath30 is the noise . in the absence of noise , for small @xmath28
the device responds as a weakly damped harmonic oscillator [ fig .
[ fig1 ] ( b ) ] . as @xmath28 increases
, the response curve bends toward higher frequencies because of spring hardening . for @xmath28 exceeding the critical value @xmath12 , the oscillator becomes bistable , with two coexisting states of forced vibrations in the range @xmath31 .
the frequencies @xmath32 depend on @xmath28 .
they correspond to saddle - node bifurcations and are indicated by arrows in fig .
[ fig1 ] ( b ) .
noise can lead to switching between the vibrational states . in the parameter range of our operation ,
thermal noise was too weak for switching to be observed .
we generated noise externally , by applying random electrostatic forces between one of the sidegates and the beam .
to create gaussian noise , @xmath33 in eq .
( [ eq : duffing ] ) , the johnson noise from a @xmath34 resistor at room temperature was amplified and directly applied to the sidegate .
generating poisson noise involves extra circuitry [ fig .
[ fig1 ] ( c ) ] . since the oscillator is most perceptive to noise at frequencies close to @xmath35 , we used noise that consisted of pulses at frequency @xmath29 with duration @xmath36 . in the rotating frame ,
as we will explain later , such pulses looked rectangular . to create them ,
the gaussian noise voltage was connected to the trigger input of a pulse generator .
whenever the input voltage , in a rare large outburst , exceeded a threshold value , the generator produced a square pulse .
the pulse width @xmath37s was much less than the reciprocal pulse rate @xmath38 , which ranged from @xmath39 ms to @xmath40 ms .
we verified that the inter - pulse intervals obeyed the poisson statistics .
the pulse signal controlled an rf switch that only turned on when a pulse was present , generating an electrostatic force with the same phase as the regular magnetomotive force .
the output of the switch is described by the random force in eq .
( [ eq : duffing ] ) of the form @xmath41 where @xmath42 is a random pulse train , with the area of each pulse @xmath43 controlled by a variable attenuator ; for chosen @xmath44 there were @xmath45 rf cycles in each pulse .
the pulses represent poisson noise with intensity @xmath46 .
we checked that a single pulse never led to switching .
to study the oscillator dynamics , we go to the rotating frame and introduce slow variables @xmath47 and @xmath48 , @xmath49 , @xmath50 , where @xmath51 and @xmath52 is the frequency detuning . functions @xmath47 and @xmath48 are the scaled in - phase and quadrature components of @xmath53 . in the rotating wave approximation ( rwa )
poisson noise ( [ eq : poisson_noise ] ) is demodulated and becomes a random force @xmath54 that drives the quadrature component @xmath48 . in dimensionless
slow " time @xmath55 eq .
( [ eq : duffing ] ) becomes @xmath56 here , dot indicates differentiation with respect to @xmath57 , and the noise @xmath58 .
the two dimensionless parameters @xmath59 and @xmath60 are the scaled intensity of the modulation and the scaled detuning of the modulation frequency from the oscillator eigenfrequency , @xmath61 functions @xmath62 do not explicitly depend on time in the rotating frame .
the stationary solutions of equations ( [ eq : eom_full ] ) in the absence of noise , which are given by equations @xmath63 , determine the stationary vibrational states of the oscillator . in the region of bistability
there are three stationary states , two of them are stable and the third is the saddle - type solution of eq .
( [ eq : eom_full ] ) in the absence of noise . for chosen pulse duration @xmath64
, the noise @xmath65 is effectively @xmath66-correlated . from eqs .
( [ eq : duffing ] ) and ( [ eq : poisson_noise ] ) , each pulse of @xmath65 shifts @xmath67 by @xmath68 .
the dimensionless pulse rate is @xmath69 and the dimensionless intensity is @xmath70 .
if the system is initially at a stable state , for small @xmath71 and @xmath72 it usually has time to relax back to this state between the pulses .
however , occasionally there may happen a sequence of frequent pulses that will push the system far from the occupied state so as to cause switching to a different state .
this switching mechanism is qualitatively different from the case of gaussian noise where individual noise pulses are too frequent to be resolved by the system and the noise outburst has to overcome the deterministic force that drives the system toward the stable state .
it is most interesting to study poisson - noise induced switching close to the saddle - node bifurcation point where , on the one hand , the switching rate @xmath0 becomes larger and , on the other hand , the system dynamics is expected to display universal features related to the critical slowing down .
the quadrature @xmath67 corresponds to the soft mode near a bifurcation point of the duffing oscillator .
the in - phase component @xmath73 follows @xmath67 adiabatically , so that @xcite @xmath74,\qquad a_b=\omega(2r_b^2 - 1),\ ] ] where @xmath75 and @xmath76 are the values of @xmath47 and @xmath48 at the bifurcation point and @xmath77 .
extending the analysis @xcite to the poisson noise , one obtains equation of motion for the soft mode as @xmath78 .
the potential @xmath79 has the form @xmath80 , where @xmath81 is the distance to the bifurcational value of @xmath48 .
the parameters @xmath82 in the expression for @xmath79 are known;@xcite @xmath6 is the distance to the bifurcation point in the space of the driving force parameters @xmath83 ; if @xmath84 , potential @xmath85 has a minimum and a maximum at @xmath86 and @xmath87 , respectively .
the minimum corresponds to the stable state of forced vibrations .
the extrema of @xmath88 merge at the bifurcation point @xmath89 and disappear for @xmath90 .
this maps the problem of the oscillator dynamics near a bifurcation point onto the problem of an overdamped particle with one dynamical variable in a static potential . using the results for the switching rate @xmath91 near a bifurcation point in a static potential,@xcite
, we obtain for the switching exponent @xmath92 @xmath93 here @xmath94 with @xmath95 .
parameters @xmath96 , @xmath97 , and @xmath98 are the bifurcational values of the scaled modulation intensity @xmath59 given by eq .
( [ eq : beta_omega ] ) , the frequency detuning @xmath99 , and the scaling factor @xmath100 , respectively ; parameter @xmath101 in eq .
( [ eq : qp_analytic ] ) is @xmath102 .
equation ( [ eq : qp_analytic ] ) applies provided the number of noise pulses @xmath103 needed to shift the system from @xmath86 to @xmath87 is @xmath104 , where the dimensionless relaxation time is @xmath105 .
we checked that this condition was met in the experiment .
intuitively , @xmath106 can be found by maximizing the probability for the poisson distribution to have @xmath107 pulses in time @xmath108 and taking the logarithm of this probability.@xcite equation ( [ eq : qp_analytic ] ) should be contrasted with the expression for the switching exponent @xmath109 in the case of gaussian noise , @xmath33 in eq .
( [ eq : duffing ] ) . for white noise , @xmath110 , near a bifurcation point@xcite @xmath111
this expression displays a different dependence on @xmath6 and the noise parameters than eq .
( [ eq : qp_analytic ] ) .
rad@xmath24s@xmath25 in the presence of poisson noise .
the amplitude is normalized to the vibration amplitude at the critical point .
the small spikes are the response of the oscillator to pulses that do not lead to transitions .
the response during reset to the low - amplitude state is not shown .
( b ) the histogram of the residence times .
the linear fit gives the mean residence time . , width=257 ] in the experiment , the driving amplitude @xmath28 was @xmath112 , so that the hysteresis loop in fig .
[ fig1 ] ( b ) was comparatively large while the vibrations remained almost sinusoidal .
we measured the switching rate near the bifurcation point @xmath113 at driving frequency @xmath114 the distance to the bifurcation point was along the frequency axis , with @xmath115 .
the oscillator was initially prepared in the low - amplitude state .
then the noise was turned on , and we monitored the oscillator amplitude .
figure [ fig2 ] ( a ) exhibits a time series of the measured amplitude in the presence of poisson noise . after switching to the high amplitude state
, we reset the oscillator to the low - amplitude state by a procedure described in ref . .
figure [ fig2 ] ( b ) shows the histogram of the residence time , demonstrating an exponential decay of the state population .
the linear fit determines the switching rate @xmath0 .
a similar behavior is observed for switching due to gaussian noise@xcite .
the measured switching rates for poisson noise were close to the theoretical values that took the prefactor into account @xcite . the dependence of the switching rate @xmath0 on the noise parameters and the distance to the bifurcation point is presented in figs .
figure [ fig3 ] ( a ) shows that , for gaussian noise , @xmath116 linearly depends on the reciprocal noise intensity @xmath117 , in agreement with eq . ( [ eq : qg_analytic ] ) .
the poisson noise intensity @xmath118 depends on both the pulse rate @xmath8 and area @xmath43 . in fig .
[ fig3 ] ( b ) we show that the increase of @xmath116 with @xmath119 for fixed @xmath43 is strongly sub - linear . replacing @xmath119 with @xmath120 gives a good linear fit , as seen from the inset in fig .
[ fig3 ] ( b ) .
this is completely different from the @xmath117 dependence for gaussian noise and is consistent with eq .
( [ eq : qp_analytic ] ) . for gaussian noise as a function of the noise intensity @xmath1 for @xmath121 rad@xmath24s@xmath25 ; @xmath1 is in the units of s@xmath122 .
the solid line is a linear fit .
( b ) the dependence of @xmath123 on the reciprocal scaled pulse rate @xmath124 at @xmath125 rad@xmath24s@xmath25 for poisson - noise induced switching ; the pulse area @xmath43 is constant .
inset : the same data plotted vs. @xmath126.,width=257 ] next , the dependence of the switching rate on the pulse area @xmath43 was studied by changing the pulse height , while keeping the pulse width @xmath44 constant . in fig . [ fig4md ] ( a ) we show @xmath116 as function of @xmath120 for different @xmath43 . fitting the data by straight lines , we found the slope @xmath127 as function of @xmath43 .
figure [ fig4md ] ( b ) shows a log - log plot of this slope .
the measured exponent of @xmath128 is @xmath129 , consistent with the theoretical value of @xmath130 in eq .
( [ eq : qp_analytic ] ) [ the factor @xmath131 in @xmath106 , eq .
( [ eq : qp_analytic ] ) , could not be measured in the studied range of @xmath43 ] . because the pulses of @xmath133 push " the quadrature @xmath134 in one direction , near a bifurcation point , where the motion in the rotating frame is overdamped , switching is possible only where this is the direction from @xmath86 to @xmath87 . for @xmath135 , eq .
( [ eq : qp_analytic ] ) describes switching near @xmath113 in fig .
switching near the second ( larger - frequency , in our case ) bifurcation point @xmath136 occurs if @xmath137 .
this was tested in the experiment by changing sign of @xmath133 in eq .
( [ eq : poisson_noise ] ) .
the switching exponent near @xmath136 is given by eq .
( [ eq : qp_analytic ] ) with @xmath138 and @xmath139 . vs. the logarithm of the reciprocal pulse rate @xmath140 at @xmath141 rad@xmath24s@xmath25 .
the solid lines are a linear fit for a fixed @xmath43 . from top to bottom , the values of @xmath43 were @xmath142 , in scaled units .
the error bars are smaller than the markers .
( b ) logarithm of the slopes from ( a ) as a function of @xmath143 .
the line is a linear fit , yielding the exponent @xmath129.,width=257 ] last but not least , we studied the dependence of the switching exponent on the distance to the bifurcation point . for gaussian noise
, we generated a number of plots similar to fig .
[ fig3 ] ( a ) , each at a different @xmath144 . we could then extract the activation barrier @xmath145 [ the switching rate @xmath146 from the slope of @xmath147 vs @xmath148 .
figure [ fig5md ] ( a ) shows @xmath149 vs @xmath150 .
a power law fit gives a critical exponent of @xmath151 in agreement with the theoretical predictions of @xmath152 in eq .
( [ eq : qg_analytic ] ) and with other experiments @xcite . for poisson pulses ,
the slopes @xmath153 were extracted from plots of @xmath116 vs. @xmath120 [ similar to the inset of fig . [ fig3 ] ( b ) ] at different @xmath144 and plotted as fig .
[ fig5md ] ( b ) . fitting the results in fig .
[ fig5md ] ( b ) with a power law @xmath154 gives a critical exponent of @xmath155 , consistent with the theoretical prediction of @xmath156 [ eq .
( [ eq : qp_analytic ] ) ] .
apart from the different exponent in the dependence on @xmath157 compared to the gaussian noise , the dependence of @xmath106 on @xmath144 for poisson pulses contains an extra log factor in eq .
( [ eq : qp_analytic ] ) .
the dynamical range of the parameters in our experiment did not allow us to single out this factor on top of the @xmath158 directly by measuring @xmath159 for different @xmath157 while staying close to the bifurcation point .
however , we made use of the fact that @xmath8 is only present inside the log factor in @xmath106 . therefore the slope @xmath153 should be independent of @xmath160 . by varying the poisson pulses via @xmath8 , the factor @xmath158 could be isolated using the slope of the activation exponent in the inset of fig .
[ fig3 ] ( b ) as function of @xmath161 for different modulation frequency .
the parameters of the oscillator used in the experiment were chosen to be close to the bifurcation values . however , they were not close enough to make the oscillator motion in the rotating frame fully controlled by one dynamical variable @xmath67 , i.e. , the assumption that the variable @xmath73 follows @xmath67 adiabatically was not strictly valid . in appendix
we show that the full theory of the escape rate in the studied range gives results very close to the asymptotic expression ( [ eq : qp_analytic ] ) .
this is because @xmath162 behaves nonadiabatically only very close to the metastable state of the oscillator , but the contribution of this area to the switching exponent @xmath106 is small . for gaussian noise induced switching ( a ) and the slope @xmath163 for poisson noise induced switching ( b ) as functions of the distance to the bifurcation point @xmath144 .
the fitted critical exponents are @xmath164 for gaussian noise in panel ( a ) and @xmath165 for poisson noise in panel ( b).,width=257 ]
our results demonstrate that driven nonlinear systems can serve as bifurcational detectors of non - gaussian noise , based on the sensitivity to the noise statistics of the dependence of @xmath159 on the distance to the bifurcation point .
the detectors can be implemented with nano - mechanical systems , rf josephson junctions , or other systems .
a stringent condition is that within the relaxation time of the system the mean number of pulses needs to remain small . as a rough estimate , to measure fluctuations in mesoscopic systems with current @xmath102 pa
, the relaxation time of the detector needs to be @xmath166 ns . a nanomechanical detector will therefore require a state - of - the - art ghz resonator with a quality factor @xmath167 . at the same time , non - gaussian noise plays an increasingly important role in mechanical devices as they are getting smaller .
besides the discreteness of the fluctuating electric charge coupled to the mechanical motion @xcite , it can come from attachment or detachment of molecules @xcite or random spin flips @xcite .
the effects of non - gaussian noise on nano - mechanical systems warrant further investigation . in summary
, we have measured the rate @xmath0 of poisson noise induced switching in a nonlinear resonator near a bifurcation point where the number of stable vibrational states changes .
the observed dependence of @xmath159 on pulse rate @xmath8 and area @xmath43 , as well as on the distance to the bifurcation point @xmath6 is consistent with the predicted scaling @xcite , which is strongly different from the scaling for gaussian - noise induced switching .
this work was supported by the nsf through grants no .
cmmi-0856374 and 0900666 .
is supported by shun hing solid state clusters lab and hkust 600312 from the research grants council of hong kong sar .
the random trajectories followed by our periodically driven micromechanical resonator in fluctuation - induced switching are concentrated around a specific path in its phase space @xmath168 , which for gaussian noise has been now seen in experiment.@xcite this path is commonly known as the most probable switching path ( mpsp ) . even though the mechanism for switching induced by poisson noise is qualitatively different from gaussian noise as discussed in the main text
, the notion of mpsp still applies . in general
, the mpsp follows complex patterns in multivariable systems .
for example , in a micromechanical resonator driven into parametric resonance , if the motion in the rotating frame is underdamped the mpsp has been shown to consist of spirals around the two attractors on the @xmath168 plane .
as the system approaches the bifurcation point , the motion near the corresponding stable state in the rotating frame undergoes a crossover from 2d to effectively 1d and becomes overdamped . in a switching event ,
the system most likely moves along a straight line on the @xmath168 plane from the initially occupied state to the saddle point ; this motion is controlled by a 1d overdamped soft mode fluctuating in a 1d potential well .
the scaling of the switching exponent predicted in ref .
is obtained using the asymptotic theory that considers the escape of the system from such potential well . here , with the parameters used in the experiment
, we analyze theoretically the motion of our system in phase space .
we show that under the experimental conditions , the portion of the mpsp that goes from the stable state to the saddle point ( the uphill " portion ) is largely linear .
spiraling only occurs close to the initially occupied attractor . for the spiraling to entirely disappear so
that the motion is truly 1d , it is necessary to choose driving frequencies even closer to the bifurcation value .
nevertheless , the spiraling portion of the path contributes only a negligible amount to the action in the calculation of the switching exponent . as a result
, the scaling behavior of the switching exponent is found to extend beyond the 1d regime , in agreement with our experimental findings . in the rotating frame , the generalized fokker - planck equation for the probability density @xmath169 of the system in the presence of poisson noise reads@xcite @xmath170+\tilde\nu_p[\rho(\mathbf{q}-\tilde{\mathbf{g}}_p)-\rho(\mathbf{q } ) ] , \label{eq : fp2d}\ ] ] where @xmath171 is the position vector in the rotating frame , @xmath172 , and @xmath173 .
the components @xmath174 are given by eq .
( [ eq : eom_full ] ) ; @xmath72 and @xmath175 are the dimensionless pulse area and pulse rate , respectively .
we are interested in the quasistationary solution of eq .
( [ eq : fp2d ] ) , which is formed over the relaxation time and persists for times small compared to the reciprocal rate of switching from the considered metastable state . respectively
, we seek the solution as @xmath176 $ ] .
we assume that on the tail of the distribution @xmath177 is large , but that it is smooth on the scale @xmath178 .
then , if we keep terms @xmath179 and assume them large , but disregard terms with higher derivatives of @xmath180 , we can reduce eq .
( [ eq : fp2d ] ) to equation @xmath181+\mathbf{p}\cdot\mathbf{k},\ ] ] where @xmath182 .
this equation maps the problem of noise - induced switching on the problem of hamiltonian dynamics of an auxiliary system .
this system has the coordinate vector @xmath183 and the momentum vector @xmath184 , its hamiltonian function is @xmath185 , whereas function @xmath180 is its action.@xcite the uphill portion of the mpsp is a heteroclinic orbit that connects the fixed points @xmath186 and @xmath187 , where @xmath188 and @xmath189 correspond to the positions of the attractor and the saddle point of the resonator in the rotating frame .
we have numerically calculated the mpsp taking into account that close , but not too close to the fixed point @xmath191 , action @xmath177 is quadratic in @xmath192 .
@xcite similar to how it is done for gaussian noise ( cf .
ref . and papers cited therein )
, one can find mpsp using the shooting method by finding the right " direction for the hamiltonian trajectories of the auxiliary system that go from the vicinity of @xmath191 , so that they approach @xmath193 .
figure [ fig : traj ] shows the uphill portion of a mpsp for typical values of the experimental parameters .
the trajectory does not show any spiraling at this length scale .
a close - up of the region close to the attractor [ fig .
[ fig : traj ] ( b ) ] reveals that spiraling indeed occurs .
therefore the uphill path is , straightly speaking , not 1d .
however , since the spiraling portion is small and lies in the region of small @xmath194 , whereas much of the uphill path remains largely straight , the asymptotic theory provides a good estimation of the switching rate . in fig .
[ fig : comparisonqp ] , we compare the measured @xmath195 with the theoretical values calculated from the the asymptotic theory for 1d dynamics [ eq . ( [ eq : qp_analytic ] ) ] and the numerical values that go beyond the 1d theory .
the black line represents the results for @xmath196 , where @xmath106 and the prefactor @xmath197 are both calculated from the asymptotic 1d theory.@xcite the good agreement indicates that the asymptotic theory provides a good estimate of the switching exponent even beyond the strictly 1d regime . the numerical results for the switching exponent @xmath106 [ eq .
( [ eq : qpnumerical ] ) ] refer to the same parameters .
since the numerical method does not yield the prefactor @xmath197 , it is not possible to perform a direct comparison between the numerically calculated @xmath106 and the experimentally measured @xmath198 . in fig .
[ fig : comparisonqp ] , , the right vertical axis ( corresponding to the numerically found @xmath106 ) is shifted from the left vertical axis ( corresponding to @xmath198 ) .
the plot indicates that the slope of @xmath106 vs. @xmath199 is in good agreement with the results from both the asymptotic theory and the measurement .
33ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) link:\doibase 10.1103/physrevlett.91.196601 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.95.176601 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.96.076605 [ * * , ( ) ] link:\doibase 10.1063/1.3430000 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevb.73.172302 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.99.060601 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | we study poisson - noise induced switching between coexisting vibrational states in driven nonlinear micromechanical resonators .
in contrast to gaussian noise induced switching , the measured logarithm of the switching rate is proportional not to the reciprocal noise intensity , but to its logarithm , for fixed pulse area .
we also find that the switching rate logarithm varies as a square root of the distance to the bifurcation point , instead of the conventional scaling with exponent 3/2 . |
the strong , spatially extended , rest - frame ultraviolet emission lines observed in high redshift radio galaxies provide one of the principal diagnostics in establishing the state of the interstellar medium in galaxies at early epochs .
the presence of a blue continuum and emission lines from regions aligned with the radio axis ( mccarthy et al .
1987 ; chambers et al .
1987 ) warned us that much of the observed optical radiation might be associated with the nuclear activity and so may not be giving us a clear picture of the stellar processes which are of such interest in studies of galaxy formation and evolution .
subsequent work has shown , indeed , that much of the blue light is scattered , polarized nuclear radiation ( e.g. , tadhunter et al .
1989 ; di serego alighieri et al . 1993 ; cimatti et al .
1993 ) and that the emission lines have a high ionization state and can not result from stellar photoionization ( mccarthy 1993 ) .
it is clearly necessary , therefore , to reach a clear understanding of the physical processes involved in the formation of the various lines and continua to be able to disentangle the stellar and the agn - related sources . for objects at high@xmath2
, the uv rest - frame lines are shifted into the optical band and the spectrum is generally dominated by ly@xmath1 , civ@xmath01549 , heii@xmath01640 and ciii]@xmath01909 . the strength of the high ionization lines suggests the presence of a hard photoionizing continuum which could originate at the agn itself ( robinson et al .
1987 ) or be associated with fast shocks generated in extranuclear regions by the radio jets ( sutherland , bicknell & dopita 1993 ) .
the strong radio / optical asymmetries observed in these objects which exhibit the ` alignment effect ' ( mccarthy et al .
1991a ) may simply result from a one - sidedness in the distribution of material .
it it is clear , however , that correlated line and continuum asymmetries could be produced by dust scattering and line fluorescence for sources where the radio axis falls significantly away from the plane of the sky . in this work ,
we concentrate on modeling the high excitation lines for which rather extreme ratios relative to ly@xmath1 have recently been reported .
the presence of dust has been universally invoked to explain the weakness of ly@xmath1 which is a resonance line and therefore , due to multiple scattering , more susceptible to absorption .
we explore the fact that any resonance line will be extremely sensitive to geometrical factors , an aspect of the problem which has so far been overlooked in modeling the uv lines .
if in radio - galaxies the distant gas clouds are photoionized from the outside by partially collimated uv radiation emitted by the nucleus , the line formation process particularly for the resonance lines is very different from internally ionized hii regions .
the escape of resonance line photons is strongly influenced by the presence of spaces between the line emitting clouds .
we have collected from the literature the observed line ratios for a number of high z radio - galaxies in which no contribution from any nuclear blr is apparent .
we have built a diagnostic diagram consisting of the lines civ@xmath01549/ly@xmath1
_ vs. _ civ@xmath01549/ciii]@xmath01909 , in which we compare the position of the objects with photoionization models which not only consider the effects of internal dust but also those of the viewing perspective the angle between the incoming ionizing radiation and the observer s line of sight .
our concentration on the particular class of radio galaxies is purely for pragmatic reasons .
it is these objects , which we presume to harbour a powerful quasar which is hidden at optical / ultraviolet wavelengths to our line of sight , which are most readily found and studied at the high redshifts where we have access to the ultraviolet spectrum from groundbased observations .
our conclusions should be equally applicable to other classes of agn . for some objects , ly@xmath1 is observed to be fainter with respect to civ than predicted by dust - free photoionization models .
the explanation previously proposed to explain the weakness of ly@xmath1 with respect h@xmath1 or h@xmath3 has been dust destruction of resonant ly@xmath1 photons .
this is _ not _ borne out by our calculations in which we have used arbitrary amounts of dust and found that this can not simultaneously weaken ly@xmath1 while leaving the civ / ciii ] ratio relatively unchanged since resonant civ suffers also from dust absorption .
alternatively , by varying the proportions of the illuminated and the shadowed cloud faces which contribute to the observed spectrum , we are better able to match the data .
geometric explanation could naturally explain that some of the brightness asymmetries noted by mccarthy et al .
( 1991a ) on sides of the nucleus .
as we find that geometry alone ( with or without internal dust ) can in principle explain most of the specific line ratios observed ( fainter ly@xmath1 compared with either civ or heii ) , we also discuss the possibility of a patchy outer halo of neutral gas to account for the diffuse ly@xmath1 seen in some cases to extend much beyond the civ emitting region and even the outermost radio lobes .
reflection by cold gas of the brighter ly@xmath1 emitting side of the ionized clouds would lead to a narrower profile for such a diffuse component .
another possibility is that part of the beamed nuclear _ continuum and blr _ radiation might be reflected at the wavelength of ly@xmath1 by thin matter - bounded photoionized gas at very large distances from the nucleus leading to a diffuse ly@xmath1 component aligned with the radio axis .
it appears to us that geometrical perspective effects are an essential component of the interpretation of the uv spectrum of radio - galaxies whether or not dust is present .
furthermore , a spectrum in which only ly@xmath1 appears does not necessarily imply starburst activity , other lines must be observed before the existence of an hii region can be inferred .
we have constructed a data sample containing galaxies at high z for which the civ , ly@xmath1 and , in most cases ciii ] , emission lines have been measured . since very high densities such as those encountered in the blr
alter significantly the line formation and transfer processes , we have excluded those objects which show evidence of a blr . in table 1 , we list the object names , the line ratios of interest to us here , the redshift and the reference to the observations .
the larger fraction of the data are taken from the recent thesis by van ojik ( 1995 ) which includes objects selected on the basis of a very steep radio spectrum . probably by virtue of the radio selection ,
these sources populate the region of the line ratio diagram ( fig .
4 ) with lower civ / ciii ] ratios ( lower ionization parameter ) than the previously published objects .
the line measurements refer to the integrated emission from the object collected with a long slit aligned with the major axis .
the data are compared to photoionization models computed using the multipurpose photoionization shock code mappings i. the version described in binette et al .
( 1993a , b ) is particularly suited to the problem since it considers both the effects of the observer s position with respect to the emitting slab and the ionizing source ( cf , fig .
we distinguish between the spectrum seen from the back and from the front of the slab .
the code also considers the effects of dust mixed with the ionized gas : extinction of the ionizing continuum and of the emission lines , scattering by the dust and heating by dust photoionization .
the treatment of the escape of resonant civ and ly@xmath1 photons in a dusty medium is described in appendix b of binette et al .
( 1993a ) and is based on the results of hummer & kunasz ( 1980 ) .
the dust content of the photoionized plasma is described by the quantity @xmath4 which is the dust - to - gas ratio of the plasma expressed in units of the solar neighbourhood dust - to - gas ratio . to specify the gas metallicity , we scale with a factor @xmath5 the solar abundance set of trace elements from anders & grevesse ( 1989 ) .
he / h is kept constant at 0.1 .
we generally consider the solar case with @xmath6 .
since the presence of dust implies depletion of refractory trace elements , we use the prescription given in appendix a of binette et al .
( 1993a , hereafter bwvm3 ) to derive the _ gaseous phase _ abundances for any metallicity @xmath5 .
the depletion algorithm is function of the ratio @xmath7 and makes use of the depletion indices listed in whittet ( 1992 ) for the different metals .
the calculations consider the gas pressure to be constant ( isobaric models ) and so the density behaviour with depth in the cloud is determined by the behaviour of the temperature and the ionization fraction of the gas .
the ionization parameter , a measure of the excitation level of the ionized gas , is defined as the quotient of the density of ionizing photons incident on the cloud and the gas density : @xmath8 where @xmath9 is the monochromatic ionizing energy flux impinging on the cloud , @xmath10 the ionization potential of h , @xmath11 the density of the gas in the front layer of the slab and @xmath12 the speed of light . [ cols="<,<,<,<,<",options="header " , ] the detailed studies of the optical emission lines of low z radio galaxies provide our basic reference for the properties of the emitting gas and the ionizing continuum ( e.g. , robinson et al .
1987 , hereafter rbft87 ) . as a guide in choosing the input parameters
, we have assumed that the excited gas of _ very _ high z radio galaxies ( @xmath13 ) has similar properties to that of the low redshift ( @xmath14 ) objects
. for the extended ( eelr ) as well as the ( narrow ) nuclear emission lines , photoionization by a hard continuum appears to best explain the various line ratio diagnostic diagrams .
the eelr follow a tight sequences which are well reproduced by varying the ionization parameter @xmath15 in photoionization models . by combining [ oi]@xmath06300/h@xmath1 with information on the gas excitation ( from the ratio [ oiii]@xmath16/h@xmath3 ) , rbft87 inferred that the ionizing spectrum is characterized by mean photon energies in the range 30 to 40 ev , which rules out normal stars .
a power law of index @xmath17 ( @xmath18 ) obeys this requirement and has produced remarkably good agreement with the optical line ratios although equally good fits can be obtained using hot ( @xmath19k ) blackbody spectra .
hereafter the above power law will be adopted as ionizing energy distribution .
concerning metallicity , rbft87 indicated that the abundances can not be much higher than solar values .
they could however be lower by a factor of a few , a distinct possibility in the case of the extranuclear gas in very high z galaxies .
unless specified otherwise , we have adopted @xmath6 in our calculations and have verified that lower abundances do not in any way affect the conclusions reached here .
the red [ sii ] doublet ratio in low z radio galaxies indicates densities for the extended emission line regions which are not higher than a few tens per @xmath20 .
we therefore adopt the low density regime , specifically @xmath21 , since the _ extranuclear _ gas generally dominates the line luminosities .
in this section we first extend the assumption of photoionization to modeling the line emission in the rest - frame ultraviolet of high z radio galaxies .
we distinguish the effects of having a back and a front view of an externally photoionized slab ( section 3.2 ) . in 3.3 ,
we introduce internal dust and discuss how , alone , it is insufficient to explain the low value of the ly@xmath1/civ ratio in some objects . in 3.4
we distinguish between the resonance line ` mirror ' intrinsic to our photoionized slab and the possibility of having external cold gas which is shadowed from the ionizing source acting as a reflector . at very high redshift ,
observational access to the normal optical plasma diagnostic lines is restricted and the available data set is small . also ,
where measurements are available , few lines are measured in any given source .
it is nevertheless interesting to compare the [ oiii]@xmath05007/h@xmath3 value observed at very high z to that of low z radio galaxies as plotted in figure 1 .
the curved line corresponds to the sequence of photoionization models of rbft87 which reproduce reasonably well the optical ( rest frame ) line ratios of the extended emission line regions and their associated narrow nuclear line emission .
the range in @xmath15 is [ @xmath22 .
the three lower horizontal lines correspond to high z radio galaxies while the upper one corresponds to the `` average '' radiogalaxy spectrum as defined by mccarthy ( 1993 ) .
the high z objects are not substantially different from the low z sample although the apparent trend towards weaker [ oiii]/h@xmath3 , if confirmed , might indicate substantially lower metallicities : low enough to overcome the higher ratios produced by the higher kinetic temperatures in moderately underabundant objects .
such an effect could also be produced by insufficient spatial resolution at high z to separate the high from the low excitation regions . to see if these models reproduce the observed civ / ly@xmath1 and civ / ciii ] ratios , we have presented the observed and predicted values in fig .
the various lines correspond to calculations in which the ionization parameter is varied , generating a sequence in @xmath15 .
different line types correspond to different viewing angles ( front , back and average ) of the photoionized slab .
the metallicity was @xmath6 and the gas dustfree .
we find that somewhat higher values of @xmath15 than in the optical are required to reproduce the civ / ciii ] ratio observed .
this is most probably a consequence of the different way in which the objects at high and low z are selected : the distant sources are all very powerful radio sources with luminous agn . for a power law of index 1.4 ,
the optimum value is @xmath23 .
hereafter , diagrams will only cover the range of [ 0.01,0.1 ] in @xmath15 .
a notable observation is the distribution of observed points above the model loci in fig . 2 even at the highest value of @xmath15 .
in the extreme objects at least like f 10214 + 4724 and tx 0211 - 122 this results from a weakening of ly@xmath1 rather than atypical values of civ / ciii ] . it is usually claimed that the destruction ly@xmath1 photons by resonance scattering in the presence of dust is the explanation for its faintness , but why does _ not _ the same process reduce civ which is also a resonance line ? is there an alternative explanation for this selective dimming of ly@xmath1 ? to answer this question , we examine the geometrical aspects of the line formation process . each emitting cloud is approximated as a plane parallel slab which contains a fully ionized region and a partially ionized zone where low ionization species co - exist ( e.g. , o@xmath24 , s@xmath25 , etc ) with a mixture of h@xmath24 and h@xmath25
. in principle there can be an additional neutral zone which does not contribute to the emission line intensities ( see fig .
3 ) . in this section
we consider the dust - free case . for most lines ( like ciii ] , heii , h@xmath3 , [ oiii ] , etc )
line opacity is negligible and the line is emitted isotropically with photons escaping freely in all directions .
however , when the line opacity is important as it is for civ and ly@xmath1 , line scattering occurs which increases the path length .
another important effect of large optical depths is that the line photon will not escape isotropically .
a resonance line photon following many scatterings must statistically escape in the direction of highest escape probability which can be shown to be the front for the photoionized slab depicted in fig .
3 . in the case of ly@xmath1 ,
the reason is that while ly@xmath1 photons are generated more or less uniformly within the slab ( except within the piz ) by recombination the neutral fraction and therefore the incremental line opacity @xmath26 increases monotonically as a function of depth as discussed in more detail by bwvm3 .
this means that for an _ open _ geometry like that shown in fig . 3
, the zone of equal escape probability of front _ vs _ back occurs far beyond the point where half the luminosity of ly@xmath1 is produced . for the collisionally excited civ line ,
the tendency to escape from the front also exists although it is less pronounced .
it arises mainly because the emissivity of civ is larger towards the front due to the temperature gradient across the c@xmath27 zone .
note that while civ is emitted and scatters within the rather limited zone containing c@xmath27 , ly@xmath1 remains subject to scattering outside the region where it is produced .
while most ly@xmath1 emission is produced in the fully ionized zone , most of the line opacity occurs within the piz .
the presence of a layer of neutral gas beyond the piz will increase the anisotropy of ly@xmath1 escape .
the effects described qualitatively above are shown in fig . 4 using detailed photoionization calculations .
we present the same sequence of dust - free models as in fig . 2 but distinguish between the spectrum seen from the _ back _ equivalent to observing the clouds through the piz from that seen from the _ front _ equivalent to seeing the uv irradiated side .
the differences are striking : both ly@xmath1 and to a lesser extent civ are fainter when seen from the back .
the ciii ] line is isotropic in the dust - free case .
the fact that ly@xmath1 is more affected by perspective is due to the significant amount of neutral hydrogen ( i.e. , large line opacity ) in the piz which acts as a mirror .
although we have considered a very simplified geometry in our calculations , the method nevertheless treats properly the essential physical effects and indicates how important the viewing direction is in this open geometry .
4 suggests that perspective effects alone ( _ without any dust _ ) are sufficient to explain the weak ly@xmath1 seen in some objects .
ionization bounded calculations with @xmath28 imply total hydrogen column densities ( h@xmath25 region + piz ) @xmath29 ( of which about 60% is ionized ) . adding a modest neutral zone beyond the piz of @xmath30 would double the civ / ly@xmath1 ratio without affecting in any way the civ / ciii ] ratio .
it seems , therefore , that a geometry where we see preferentially the ionized gas from the side of the piz gives us an explanation of the weakness of ly@xmath1 .
how would this apply to the eelr of the powerful radio galaxies ? in a very simplified scheme , we can imagine ( see fig .
5 ) that the clouds seen from the nearside cone are seen from a direction which we approximate as the back perspective in our slab calculations while clouds on the far side would be seen from the front .
the studies of mccarthy et al .
( 1991a ) which emphasized the one - sideness of the line brightness suggest that the observer with limited spatial resolution at very high z will be biased towards either a back or front dominated perspective depending on whether it is the near or the farside illuminated cone which is intrinsically brighter . our proposed interpretation of the high civ / ly@xmath1 ratio of some objects in fig .
4 is that they correspond to the case where the brightest clouds are seen from behind .
note that , because of the intensity weighting , there is rather little difference between the pure front perspective and the sum of equal numbers of front- and back - clouds , a small effect which is not easily distinguished from that of slightly reducing the value of @xmath15 .
our conclusion from this section is that perspective effects in an open externally illuminated cloud geometry go a long way towards explaining the behaviour of the civ , ly@xmath1 , ciii ] line ratio diagram .
it implies that , when spatially resolved spectra become available , marked asymmetries in the civ / ly@xmath1 ratio could arise when the axis of the illuminated cones forms a large angle with the sky plane .
there is evidence for the existence of dust in some high z radio galaxies .
the detection of 4c41.17 ( @xmath31 ) and b2 0902 + 34 ( @xmath32 ) ( chini & krgel 1994 ; dunlop et al . 1994 ) and 8c1435 + 635 ( @xmath33 ) ( ivison 1995 ) in the mm spectral range is attributed to warm dust .
also , the iras galaxy f10214 + 4724 ( @xmath34 ) has been shown , from the far infrared flux , to contain @xmath35 of dust , although this figure may be reduced by the gravitational lensing amplification factor ( serjeant et al .
in addition , the mid- to far - infrared measurements at intermediate z are explained as emission from warm dust ( heckman , chambers & postman 1992 ) .
we should emphasize , however , that we have no direct measurement of how this dust is spatially distributed . in the event that this infrared emission arises from the reprocessing of higher energy photons from the agn by a dusty torus
, it has no direct bearing on our modeling of ionized gas at tens of kpc . however , there is considerable evidence that aligned blue polarized continuum is the result of scattering of the anisotropic nuclear radiation field by dust ( e.g. , cimatti et al .
in this case it is very likely that at least some of the dust is internal to the extended line emission regions . in this section
we consider the effects of including dust within the gas clouds which emit the uv lines .
we illustrate the effects of dust mixed with the emitting gas in fig . 6 where we plot sequence of models which correspond to the front perspective for three different dust - to - gas ratios : @xmath36 , 0.3 and 1 .
the effects on the line ratios are evident .
the resonance scattering suffered by civ and ly@xmath1 increases their pathlengths many times and , therefore , the probability of their being absorbed by dust grains is much higher than for ciii ] . in the case of ly@xmath1 , however , the geometrical thickness of the h@xmath25 region exceeds greatly that of the c@xmath27 since we observe more than one stage of ionization of metals ( e.g. c@xmath37 ) .
any reasonable parameters for the ionization structure of a photoionized slab with @xmath38 indicates that the opacity in ly@xmath1 greatly exceeds that of civ , which implies a larger pathlength increase for ly@xmath1 than for civ .
this results in relatively more ly@xmath1 absorption by dust .
this effect explains how the ratio civ / ly@xmath1 increases somewhat with increasing @xmath4 .
dust absorption of resonant civ on the other hand causes a comparable decrease in civ / ciii ] .
what is important in these results is that , even with a concentration of internal dust as high as @xmath39 ( equivalent to that in solar neighbourhood cold clouds ) , it is not possible to reproduce the high ratio civ / ly@xmath40 which is observed in some objects and has been attributed to dust .
we emphasize that higher amounts of dust do not change these results .
for instance models with @xmath41 do not lead to any higher ly@xmath1/civ ratio than the models shown in fig .
when the clouds are viewed from behind , even with @xmath42 they are sufficiently opaque that all the uv lines are severely absorbed . with @xmath43 and @xmath39 ,
all of the high excitation uv lines become absorbed within the piz .
to produce an acceptable spectrum , without reddened civ / ciii ] and civ / heii ratios , as seen from the back of an ionization bounded slab with @xmath28 , we have to use much smaller amounts of dust like @xmath44 ( 2% of local ism ) in models . in fig . 7
, we plot the back and front sequences for such models .
these can reproduce the weak ly@xmath1 objects although this result is obtained only for the back spectrum , emphasizing that perspective is the dominant factor .
it should be noted that the amount of extinction within the ionization bounded slab implied by @xmath45 ( i.e. , @xmath460.2 ) is consistent with the quantity of small dust grains needed to explain the extremely blue continuum of the `` detached '' ionized cloud in pks 2152 - 69 ( di serego alighieri et al .
1988 ; magris & binette private communication ) , supposing that the continuum energy distribution is the result of dust scattering of the nuclear radiation . note that because the scattering / absorbing dust is locally kinematically linked to the emission line gas , we require much smaller column densities of hi than the absorbing screen proposed by van ojik et al .
( 1994 ) ( @xmath47 ) to explain civ / ly@xmath48 .
the line widths and centroids of the piz and of the fully ionized gas are expected to be quite similar within each of the emitting clouds , the ensemble of which could have a greater ` turbulent ' velocity dispersion . to reproduce the extreme case of the iras galaxy f10214 + 4724 with civ / ly@xmath49
, we need additional neutral gas beyond the piz .
for instance , it requires only a column density of @xmath50 assuming @xmath51 .
alternatively we might consider that @xmath4 within the piz increases with depth as the degree of ionization decreases .
in this case , no additional hi gas would be required . if this iras galaxy is indeed an extreme seyfert 2 as argued by elston et al .
( 1994 ) , then a closed , dust enshrouded geometry as proposed by bwvm3 to explain the lyman and balmer decrements in seyfert 2 would be more appropriate than the open geometry adopted here which may be applicable only to the truly extended large scale gas .
it is likely that objects with unusually strong nv@xmath52 emission ( such as in f10214 + 4724 and tx0211 - 122 ) are cases where the nv originates predominantly from the inner nlr .
a high nv / civ ratio indicates very enriched gas which is not unexpected within the inner parts of an agn ( hamman & ferland 1993 ) and is consistent with the lack of convincing evidence that nv is spatially resolved in any of these objects .
so far we have considered the effects of scattering by gas which forms part of the line emitting clouds : we refer to this as an ` intrinsic ' process .
similar effects could be produced by a large - scale distribution of predominantly neutral material surrounding the emitting regions : we will refer to this as the ` extrinsic ' case .
any extrinsic neutral gas component with a non - negligible covering factor could affect the observed spectra in a way which mimics that of the back perspective described earlier .
let us suppose that this outer material is broken up into cold gas clumps which are randomly distributed .
in such a case , some ly@xmath1 photons which leave the ionized cones will escape the region through the holes between the external clumps while others will strike the neutral clumps and be immediately scattered away to escape eventually through another hole in a different direction ( see fig .
if observed with sufficient spatial resolution , such a geometry would result in holes in the ly@xmath1 brightness due to reflection by intervening clumps as well as diffuse sources corresponding to reflection from clumps on the far side of the source .
the bulk of the ly@xmath1 luminosity would be preserved but redistributed on an apparently larger scale than the true line emitting clouds .
only a more closed geometry would result in a significant destruction of ly@xmath1 , assuming that the interclump space does not contain pure dust segregated from the gas phase . the reflection efficiency of clouds will , of course , depend strongly on the relative velocity fields of the emitting and the cold regions .
if the extended gas has a large scale ordered motion but small ` microturbulence ' , the reflection effects would be localized .
broad ly@xmath1 emission and continuum from the agn could , however , be scattered by any of the extranuclear clouds ( see sect .
3.4.2 ) . in general , we would expect the diffuse , scattered ly@xmath1 to show a narrower line than the integrated emission profile .
we will review the observational evidence for the existence of such large scale mirrors .
pks2104 - 242 ( mccarthy et al .
1990b ) shows very extended emission lines of ly@xmath1 , civ and heii associated with continuum knots which lie between the radio lobes and are aligned with the radio axis .
the ly@xmath1 image shows emission resolved into three distinct clumps , two of them corresponding roughly with the two continuum knots . in addition , there appears to be a low surface brightness halo of emission surrounding the entire object .
is this halo a consequence of reflection by neutral material of either ly@xmath1 emission or of the intense nuclear _ continuum _ , or are these photons emitted locally by h@xmath25 recombination ?
a way of discriminating between these two possibilities is to look for the detection of any other emission line in the halo .
if ly@xmath1 is the result of reflection by cold hi , there will be _ no _ other lines ( except possibly resonant mgii ) .
if it is instead produced by recombination , other lines should be detected in the rest - frame optical band like h@xmath1 , [ oiii]@xmath05007 , etc . from images of this object , mccarthy et al .
( 1990a ) report the existence of ly@xmath1 emission extending over a region covering 170kpc .
the ly@xmath1 emission is elongated and well aligned with the inner radio - source .
an intriguing aspect of the observations is the one - sidedness of the civ emission ( obtained with an long slit aligned with the axis of elongation ) .
the decreasing linewidth of ly@xmath1 on the side where civ is absent ( south ) suggests to us the possibility that ly@xmath1 in the south corresponds to scattered _ continuum and blr ly@xmath1 photons _ by hi clumps lying along this direction provided the ionizing radiation has already been reprocessed ( filtered out ) within the nuclear regions into nlr or blr emission . failing that ( scattering by hi clumps )
, an alternative possibility resides in very thin matter - bounded photoionized sheets of gas which can be very efficient at reflecting the intense ( beamed ) nuclear continuum as well as blr ly@xmath1 photons . due to its large scattering cross section , very small column depths of h@xmath24 ( a trace specie within the ionized phase )
are sufficient to scatter effectively the impinging flux within the core of the ly@xmath1 thermal profile .
if we suppose the existence within the cone of ionization of the radio - galaxy of a population of clouds of similar physical conditions to that thought to apply to ly@xmath1 forest clouds , the energy reflected due to resonant scattering by hi typically exceeds that generated within the clouds by reprocessing of the ionizing radiation ( i.e. by recombination ) . in effect , adopting @xmath53 as a typical column density of such cloud , we derive an equivalent width in absorption ( i.e. the scattered intensity ) of @xmath54 which corresponds to the saturated part of the ly@xmath1 curve of growth , assuming @xmath55kms@xmath56 . adopting for definiteness @xmath28 and @xmath57
, we obtain that the fraction of reprocessed ionizing photons , @xmath58 , within a very thin photoionized sheet . ]
is given by @xmath59 the equivalent width of ly@xmath1 in emission as a result of recombination is simply @xmath60 where @xmath61 is the covering factor .
if we integrate the absorbed intensity over the same area as that given by @xmath61 , the ratio of scattered versus emitted ly@xmath1 is @xmath62 ( the ratio between the above two equivalent widths ) . allowing for the fact that the cloud will be exposed not only to the nuclear continuum but to the blr ly@xmath1 which typically peaks at twice the continuum level , the ratio can be expected to increase up to 8 times the ly@xmath1 emitted by recombination . to account for the observed fwhm @xmath63 kms@xmath56 in 3c294 ( south ) , one most rely on a turbulent velocity field for the the ly@xmath1 clouds .
we ought to consider seriously the possibility that ly@xmath1 on the southern knot of 3c294 corresponds to large scale scattered light by either hi or even ionized gas unless other bona fide emission lines were detected .
so far , no other lines than ly@xmath1 are observed which is consistent with our suggestion . the only certain way of discriminating between reflection and _ in situ _ emission would be the detection at this radius of _ any _ non - resonance line in either the optical ( [ oii ] , [ oiii ] , h@xmath3 etc . ) or the uv ( cii ] , ciii ] ) . to conclude , we believe that hii regions or starbursts should not be considered the default emission mechanism in cases where only ly@xmath1 is present , especially along the axis where the agn supposedly collimates its intense nuclear continuum + ly@xmath1(blr ) light . the radio galaxy 0943 - 242 shows a ly@xmath1 profile with several absorption features which cover the whole of spatial extent of the ly@xmath1 emission ( rttgering et al .
the strongest of the absorbers is at least as extended spatially as the emitting region ( @xmath64kpc ) and its hi column density is @xmath65 .
the absorption line is blue - shifted by 250 km s@xmath56 with respect to the emission peak .
the ` screen ' is kinematically distinct from the emitting gas and therefore clearly external to the emitting clouds ) .
this is a clear demonstration that hi screens or mirrors depending on the perspective exist on galaxy scales at these early epochs . from fig . 5 of rttgering et al .
, we estimate that this cloud would reflect @xmath66% of the ly@xmath1 .
thicker clouds could exist around other objects but would be hard to detect when only the faint wings at the extremities of the emission profile were transmitted . in the case of 0943 - 242 ,
even if the screen contained dust , its effects would be negligible .
the reason , as stated before , is that ly@xmath1 photons ( seen by the cloud ) are incident from the exterior and will be very effectively scattered away before being absorbed by the dust .
even the photons getting through ( i.e. , without any scattering ) to us because they are sufficiently far in the wings of the profile would not see much dust in this particular cloud . with such low column densities , @xmath67 for @xmath68 and 0.005 for @xmath69 .
much thicker clouds may result in some non - negligible extinction but again this does not arise because of the multiple scattering of ly@xmath1 .
if the large - scale extended emission line regions ( eelr ) in radio galaxies are photoionized predominantly by the collimated uv radiation emitted by a hidden agn , the line emission and transfer processes are characterized by what we call an open geometry with externally illuminated clouds .
this produces an emission line spectrum which is distinctly different from an internally ionized hii region , especially in the ultraviolet .
our principal conclusion is that internal dust , even in proportions as large as that found in the solar neighbourhood , does not satisfactorily explain the large civ / ly@xmath1 ratios commonly seen in groundbased observations of high redshift radio galaxies .
note that the maximum amount of dust considered here ( @xmath70 ) should be considered a sensible upper limit on the ground that : a ) the extended large scale gas ( @xmath71kpc ) in radio galaxies is likely to be less than solar in metallicity , b ) the physical conditions found within the higly ionized warm plasma threatens dust survival and a more realistic approach would favour using @xmath72 . in any event , more dust does not lead to any higher ly@xmath1/civ ratio as we have verified with models in which @xmath41 .
we have demonstrated that geometrical effects can , in most cases without invoking any dust , explain the observed trends in the data .
although we agree that internal dust can quite effectively kill the ly@xmath1 luminosity , we consider essential to examine also its effects on the other uv lines especially civ which is also a resonance transition .
although some internal dust within the ionized gas is allowed by our modeling and even improve things in a few cases , we find that the effects of geometrical perspective _ dominate _ the intended decrease of ly@xmath1 without affecting excessively the other lines .
the current work demonstrates that the faintness of ly@xmath1 is not sufficient per se to imply the existence of large amount of dust in these objects .
the influence of the geometrical factors discussed above on the line spectrum gives us , potentially , a natural explanation for some of the emission line asymmetries seen in the alignment effect in radio galaxies .
we have also investigated the effects that material external to the ionized cones will have on the appearance of the object in the light of ly@xmath1 .
we suggest that the existence of the diffuse ly@xmath1 halos observed in some high z radio galaxies could be due to the reflection of ly@xmath1 photons by neutral clumps lying outside the ionization cones .
in addition to reflecting ly@xmath1 , such neutral clouds could be responsible for the spatially extended absorption seen within the ly@xmath1 emission profile .
pure ly@xmath1 emission clouds need not , therefore , necessarily be identified as star - forming regions .
mv - m acknowledges support from the deustche forschungsgemeinschaft ; also thanks the observatoire de lyon for its hospitality during a collaborative trip in june 1995 and jacco van loon for useful discussions . | in the powerful , high redshift radio galaxies , it is believed that the dominant source of ionization for the interstellar gas is the hard radiation field associated with the active nucleus .
the photon source is generally external to the clouds being ionized and so the geometrical perspective from which the gas is observed and the presence and distribution of dust must be properly accounted for in the diagnostic process . in this paper , we examine the formation of the three strong lines : civ@xmath01549 , ly@xmath1 and ciii]@xmath01909 which are often observed in the nuclear and extended emission from these sources .
we find that the observed trends , in particular the high civ@xmath01549/ly@xmath1 ratio , are often better explained by geometrical ( viewing angle ) effects than by the presence of large quantities of dust either within or outside the excited clouds .
we show that neutral condensations along the line - of - sight , by reflecting photons near the wavelength of ly@xmath1 , can increase the observed civ / ly@xmath1 ratio . the existence of hi absorption clouds ( i.e. , mirrors ) external to the emission region leads also to the presence of large , diffuse haloes of what appears to be pure , narrow ly@xmath1 emission . atomic processes
ism : dust galaxies : cooling flows , radio galaxies , galaxy formation |
recent x - ray observations have confirmed that collimated jets can be found emanating from pulsar engines ( weisskopf et al .
2000 ; helfand et al . 2001 ; gaensler et al . 2002 ; slane 2002 ; pavlov et al . 2003 ) .
these jets morphologically resemble those of young stellar objects ( reipurth & bally 2001 ) , planetary nebulae , ( balick & frank 2002 ) active galactic nuclei ( ferrari 1998 ) and microquasars ( mirabel 2001 ) . jets and disks
are likely associated in these systems , even if the role of the disk vs. central object is not fully understood with respect to jet production . here
we explore the possibility that there may also be disks around jetted pulsars , and that winds from these disks might collimate the pulsar wind .
the motivation is that relativistic pulsar winds do not seem to self - collimate easily ( e.g. lyubarsky & eichler 2001 ; bogovalov & khangoulian 2003 ; komissarov & lyubarsky 2003 ) : strong outward electric forces compete with the self - collimating collimating hoop stresses .
tsinganos & bogovalov ( 2002 ) simulated the generic efficacy of relativistic wind collimation by an ambient non - relativistic wind , but did not dynamically relate the external wind to the properties of an underlying disk .
michel & dessler ( 1981 ) suggested that disks formed from the stellar collapse offer a physical similarity between radio and x - ray pulsars .
such disks can also influence pulsar braking indices and timing ages : the former often deviate from those expected if spin - down were due only to dipole radiation , and the latter often differ from inferred actual ages .
various intrinsic mechanisms for these deviations have been proposed ( macy 1974 ; manchester & taylor 1977 ; chanmugam & sang 1989 ; beskin et al .
( 1993 ) ; chubarian et al .
2000 ; wu et al . 2003 ) , including an external torque provided by a fallback accretion disk ( e.g. menou et al .
2001a ; marsden et al .
2001 ; alpar et al .
2001 ; qiao et al . 2003 ; shi & xu 2003 ) .
this would imply a weaker pulsar magnetic field than that inferred from the measured period @xmath0 and its derivative @xmath1 .
this has been confirmed for one neutron star where the magnetic field was measured from cyclotron lines ( bignami et al .
2003 ) . here
we calculate the effect of a wind - emitting accretion disk on the pulsar spin evolution by coupling the latter to evolution equations for accretion and radius dependent mass loss .
we constrain the pulsar magnetic field strength and the initial outer accretion rate ( or disk mass ) for which the solutions match the observed periods , timing ages , and braking indices of crab and vela .
the resulting solutions produce sufficient disk - wind momenta to potentially collimate and help illuminate the pulsar winds .
we describe our wind and accretion parameterization in section 2 . in sec .
3 we show that the launch mechanism determines the radial dependence of the mass outflow rate . in sec .
4 we calculate the spin evolution of the pulsar and find best fit solutions for crab and vela .
we compare the pulsar and disk - wind momenta in sec . 5 , and conclude in sec . 6 .
given the accretion rate at the outer disk radius @xmath2 , the time - dependent mass loss from a disk - wind launched between radii @xmath3 and @xmath4 from the pulsar can be parameterized as _
dw(r , t ) = _ out(t ) .
[ eq : mwind ] ( values of @xmath5 and the parameter @xmath6 will be discussed below . )
the accretion rate at the inner edge of the disk where the torque is exerted , is given by @xmath7 .
the disk is allowed to penetrate the magnetosphere up to the magnetospheric radius @xmath8 ( @xmath9 g is the pulsar s magnetic field and @xmath10 its mass ) where the torque exerted by the magnetic field on the disk is of order the viscous torque . to constrain @xmath11 , we appeal to previous work .
chevalier ( 1989 ) studied fallback accretion , but without angular momentum .
here we adopt the menou et al .
( 2001a ; see also chatterjee et al .
2000 ) analogy between fallback accretion and accretion of the torus of gas formed by a disrupted star ( cannizzo et al .
cannizzo et al .
( 1990 ) showed that there is a transient spreading phase of duration @xmath12 during which the accretion is nearly constant , @xmath13 but after which the accretion rate declines as a power - law with time , @xmath14^{-\delta}$ ] , and the initial accretion rate depends linearly on the product of the viscosity parameter @xmath15 ( shakura & sunyaev 1973 ) and the initial mass of the disk .
they also found that @xmath16 provides a good fit to the late time evolution and that the outer edge of the disk then evolves as @xmath17 .
the initial @xmath18 is determined by the angular momentum of the presupernova star and the amount of ejected material .
menou et al .
( 2001b ) estimate @xmath19 cm for the typical angular momentum in the presupernova star found in numerical simulations ( heger , langer & woosley 2000 ) . the time @xmath12 satisfies @xmath20 the viscous infall time at @xmath21 , where @xmath22 is the outer disk height . from ( [ eq : mwind ] ) ,
we can also write the total mechanical disk - wind luminosity as l_dw(t ) = 12 _ r_in^r_out ddr ( _ dw(r , t ) v_dw^2 ) dr [ eq : lwind ] where @xmath23 is the escape speed at the inner radius , and the last step holds when the total mass inside radius @xmath24 satisfies @xmath25 .
note that the initial total disk mass is @xmath26 , so our equations could be scaled in terms of this quantity .
for magnetically mediated outflows ( e.g. spruit 1996 ) , the mechanical wind luminosity is powered by poynting flux from large scale fields .
the associated magnetic luminosity is l_dw = c4^r_out_r_in ( ) 2r dr= 12^r_out_r_in v__z _
r dr , [ poyn ] where @xmath28 is the mean surface electric field , @xmath29 and @xmath30 are the surface components of the mean magnetic field , and @xmath31 is the keplerian speed . the surface magnetic fields in ( [ poyn ] )
can be related to the fields in the disk by using the volume integral @xmath32 taken over a wedge in the disk and converting to a surface integral . using this
, tan & blackman ( 2003 ) showed that if large scale disk fields are obtained from a helical dynamo , driven by disk turbulence and shear , the surface magnetic stress takes the form @xmath33 where @xmath34 is the mid - plane sound speed , @xmath35 is the disk height and @xmath36 is the mid - plane density .
( the proportionality to @xmath37 would arise for any helical or non - helical dynamo , but the @xmath38 factor would be replaced by @xmath15 if we used the averaged turbulent magnetic stress rather than the magnetic stress associated with the large scale fields . ) combining this magnetic stress with @xmath39 , the equation for @xmath40 , the disk scalings ( e.g. frank et al .
2002 ) @xmath41 , and plugging the result into ( [ poyn ] ) gives ^r_out_r_in _ ss^1/2m_inv_^2 rdr = ^1/2_ssgm_nsm_out2 r_i ( r_out / r_in)^p ^_o_1^p-2d . [ 12 ] the last equation follows from the expression for @xmath42 , @xmath43 , and defining @xmath44 and @xmath45 . we can set ( [ 12 ] ) equal to ( [ eq : lwind ] ) , to obtain ( 1-p)(_o^p-1 ) = ( 1-p)(v_escv_dw)^2 ^1/2_ss2 ^_o_1^p-2d= ( v_escv_dw)^2 _ ss^1/22(1-_0^p-1 ) .
[ 13 ] eqn .
( [ 13 ] ) shows that @xmath27 is not independent of @xmath46 and @xmath15 once a launch mechanism is chosen .
the above formalism involves the relation that the mechanical disk - wind luminosity at a given radius is a constant fraction of the accretion luminosity , so ref ( [ 12 ] ) implies that both @xmath47 and the accretion luminosity increase with decreasing @xmath27 if all other quantities were fixed .
this formalism breaks down when the value of @xmath27 is so low that @xmath48 .
the pulsar dipole luminosity for a neutron star of angular speed @xmath49 and radius @xmath50 is given by l_pul(t ) , [ eq : lpuls ] where @xmath51 , and @xmath52 is the angle between the magnetic and spin axes .
we have not considered the time evolution of @xmath53 here .
this was studied by beskin et al .
( 1993 ) and could be included in future generalizations .
the rotational energy evolution is determined by this dipole radiation and a model dependent accretion torque so that = i = - ^4 + 2_inr_m^2_k(r_m)^ , [ eq : spindown ] where @xmath54 is the moment of inertia of the star , @xmath55 , and we use a @xmath56 model .
we also define the braking index and timing age as @xmath57 and @xmath58 respectively .
we determine the spin evolution of the star by numerically solving eq .
( [ eq : spindown ] ) combined with the equations in section 2 . to do so
, we also need the pulsar s magnetic field , the initial spin period , and the initial accretion rate in the disk .
if the star were spinning down only by dipole radiation , then its magnetic field would be directly determined from measuring @xmath0 and @xmath1 .
but when a disk torque contributes to the spin - down , we only know that the magnetic field strength must be smaller than the value obtained for a pure dipole . for young pulsars
, we expect @xmath59 g. the initial spin period @xmath60 is also uncertain , but typically , @xmath61 msec .
the initial accretion rate is likewise unknown .
to constrain these parameters in our model , we use observational values of the period , timing age and braking index for the specific pulsar under consideration .
we then solve the coupled system of equations using a grid of values for @xmath60 , @xmath62 and @xmath63 to obtain a best fit parameter range that reproduces the observed values of @xmath0 , @xmath64 and @xmath65 to within a few percent .
we specifically consider the crab and vela pulsars . in 1972 ,
the crab s true age was 918 yr , and its measured timing age was @xmath66 yr , while the braking index was @xmath67 with period @xmath68 ms .
vela has a braking index of @xmath69 ( lyne et al .
1996 ) , a period @xmath70 ms , and a timing age @xmath71 kyr , for a pulsar age ( determined from the associated supernova remnant ; aschenbach , egger & trumper 1995 ) in the range 18 - 30 kyr . here
we use @xmath72 kyr .
1a shows two solutions ( @xmath73 and @xmath74 ) for the crab , which reproduce the measured values of @xmath0 , @xmath64 and @xmath65 .
the @xmath73 solution was obtained with @xmath75 g , @xmath76 msec , and @xmath77 g / s , while that with @xmath74 has @xmath78 g , @xmath79 msec , and @xmath80 g / s .
the bottom panel displays the pulsar and wind luminosities [ eq.([eq : lpuls ] ) and eq.([eq : lwind ] ) ] during the pulsar lifetime .
1b shows , for the same values of @xmath27 , the solutions for vela which reproduce its observations within the measurement uncertainties .
the solution with @xmath73 was obtained with @xmath81 g , @xmath82 msec , and @xmath83 g / s , while that with @xmath74 has @xmath84 g , @xmath82 msec , and @xmath85 g / s with @xmath86 from our fits gives initial disk masses @xmath87
. the values of @xmath86 in our formal solutions , like those in accretion models of gamma - ray bursts , are highly super - eddington .
such disks would be dense , hot , and thick from radiation trapping ( di matteo et al .
2002 ) . here
the cannizzo et al .
( 1990 ) similarity solutions may be inapplicable , but this does not strongly affect the present day pulsar parameter solutions . ] also displayed are the pulsar and wind luminosities , and the observed lower limit on the jet luminosity .
pavlov et al . ( 2003 ) estimated an energy injection rate @xmath88 g / s for the vela jet . in all of our solutions ,
the corotation radius is less than the magnetospheric radius at all times , so the pulsars are always spun down by the disk .
the precise value of @xmath27 is not crucial to obtain a good solution for @xmath0 , @xmath64 and @xmath65 ; what matters for the spin - down torque is the value of @xmath89 , and a different @xmath27 can be compensated with a different @xmath90 to yield a similar @xmath89 ( and hence a similar evolutionary track ) .
this is why the two lines in the top three panels of fig.1 overlap .
note that the higher @xmath27 curves for @xmath47 are @xmath91 the lower @xmath27 curves because @xmath2 is larger for the larger @xmath27 cases , and this counteracts the dependence of @xmath47 on @xmath27 discussed below eq .
( [ 13 ] ) . the particular solutions in fig .
1 are not unique ; multiple combinations of the parameters @xmath92 , @xmath60 and @xmath93 can reproduce the values of @xmath0 , @xmath64 and @xmath65 within the observational uncertainties . however , we have shown solutions that also produce a disk - wind powerful enough to influence the jets observed in these sources , as discussed in more detail in the next section . for the solutions of fig . 1 , @xmath94 cm and @xmath95 cm for crab and @xmath96 cm and @xmath97 cm for vela . for the crab ,
the predicted optical emission from a even a face - on thin disk would be below the optical detection ( sollerman 2003 ) .
for vela , the observed optical flux limits require either ( i ) a disk almost edge - on to the line of sight ( ii ) disruption of the outer parts of the disk ( maybe through interaction with the supernova remnant ) or ( iii ) a geometrically thick and optically thin disk , like an advection dominated accretion flow ( adaf ) ( e.g. esin et al .
relativistic poynting flux pulsar winds do not easily self - collimate ( e.g. lyubarsky & eichler 2001 ) ; as strong outward electric forces compete with the collimating magnetic hoop stress .
( this problem would not arise in non - relativistic outflows , as the electric field @xmath98 magnetic field . )
we suggest instead that pulsar jets may be collimated by a surrounding disk wind .
a necessary condition for the disk wind to collimate the pulsar wind is that the ram pressure associated with the disk wind exceed the magnetic pressure associated with the poynting flux dominated pulsar wind on scales at or less than those where the collimation is observed .
since the jets appear on scales @xmath99 than that of the disk , it is reasonable to make this comparison far outside the pulsar s light cylinder .
magnetic energy dominates the particle energy in the pulsar wind , so the energy outflow rate can be estimated as @xmath100 , where @xmath101 is a spherical radius . comparing the two pressures amounts to comparing @xmath102 to @xmath103 .
but @xmath104 , so this amounts to comparing @xmath105 with @xmath106 , namely comparing the rate of change of the disk - wind momentum to that of the pulsar wind .
2 shows precisely this comparison for crab and vela .
the solutions indicate that a collimating influence of the disk - wind on the pulsar wind is possible for the @xmath73 case for both pulsars . that @xmath73 provides higher disk - wind momenta and higher @xmath107 than @xmath74 results because the @xmath2 which best fits the spin parameters is larger for the large @xmath27 cases .
( for the same @xmath2 , the large @xmath27 case would have a lower @xmath47 and momentum than the low @xmath27 case , as discussed below eq .
( [ 13 ] ) . )
although this effect dominates , a competing effect is that @xmath108 increases with decreasing @xmath27 : for both pulsars , eq .
( [ 13 ] ) for @xmath109 reveals that @xmath73 and @xmath74 correspond to @xmath110 and @xmath111 respectively .
collimation also requires that the vector sum of the pulsar wind and disk wind momenta incurs a significant vertical component .
even if the disk wind were spherical , when its ram pressure exceeds that of the pulsar wind the pulsar wind would incur some collimation .
however , since the disk wind is largely non - relativistic , magnetic self - collimation of the disk wind is not as problematic as that of the pulsar wind .
self - collimation of the disk wind would favor even more collimation of the pulsar wind . a more detailed model of the interacting outflows
is required to further develop these ideas .
note that tsinganos & bogovalov ( 2002 ) have found numerically that relativistic central outflows could in principle be collimated by non - relativistic disk winds , also motivating more detailed models .
a standard limitation of such numerical calculations is the assumption of the disk field geometry as a given initial condition . at the interface of the two winds ,
significant dissipation can occur , and accelerated particles can produce the observed radiation .
in addition , bending of the collimated jet could result from mutual non - axisymmetry of coaxial winds , or from misaligned pulsar and disk wind symmetry axes .
pavlov et al . ( 2003 ) suggested that a hidden wind from within the supernova remnant is needed to account for vela s bent jet .
winds from fallback accretion disks below presently detectable limits may deposit enough momenta to collimate and bend the crab and vela pulsar winds and account for their observed jets .
these disks can also account for the observed periods , braking indices , and timing ages of the crab and vela pulsars through the action of the disk torque on pulsar spin - down .
our results motivate further efforts to detect disks around jetted pulsars , and more detailed models of the nested interacting winds . collimating a relativistic central wind by a surrounding accretion disk - wind
may also be a relevant paradigm for active galactic nuclei and gamma - ray bursts .
fig . 1 : evolution of the period , braking index , timing age , dipole luminosity and disk - wind luminosity for crab and vela .
the points mark the observed values , though no such constraints exist for the crab jet power .
the age for vela is taken to be that of the associated supernova remnant .
boldface lines correspond to @xmath73 and thin lines to @xmath74 .
results are rather insensitive to the value of @xmath27 . to | for many astrophysical sources with jets , there is evidence for the contemporaneous presence of disks . in contrast , pulsars such as the crab and vela show jets but have not yet revealed direct evidence for accretion disks .
here we show that for such pulsars , an accretion disk radiating below detectable thresholds may simultaneously account for ( 1 ) observed deviations in the braking indices from that of the simple dipole , ( 2 ) observed pulsar timing ages , and ( 3 ) possibly even the jet morphology via a disk outflow that interacts with the pulsar wind within , collimating and/or redirecting it .
epsf # 1([#1 ] ) # 1([#1 ] ) # 1#1 # 1\ { # 1 } # 1 | ( apj lett . , in press ) : ( stars : ) pulsars : general ; ( stars : ) pulsars : individual ( crab , vela ) ; accretion , accretion disks ; ism : jets and outflows ; x - rays : ism ; plasmas |
there exist many interesting systems in high energy physics whose action contains an imaginary part , such as qcd at finite baryon density , chern - simons theories , systems with topological terms ( like the @xmath0-term in qcd ) and systems with chiral fermions .
this imposes a severe technical problem in the simulations , requiring an exponentially large amount of data for statistically significant measurements as the system size is increased or the critical point is approached .
furthermore , the overlap problem appears when standard reweighting techniques are applied in such systems and it becomes exponentially hard with system size to visit the relevant part of the configuration space . in @xcite
it was proposed to take advantage of a factorization property of the distribution functions of the observables one is interested to measure .
this approach can in principle be applied to _ any _ system and it eliminates the overlap problem completely .
in some cases it is possible to use finite size scaling to extrapolate successfully to large system sizes where it would have been impossible to measure oscillating factors directly . the method has been applied successfully in matrix models of non perturbative string theory ( ikkt ) @xcite , random matrix theory of finite density qcd ( rmt ) @xcite as well as the 2d @xmath1 model , the 1d antiferromagnetic model with imaginary @xmath2 and the 2d compact u(1 ) with topological charge @xcite . in this paper
we present our results @xcite for ikkt and rmt . in the first case we study the space
time dimensionality hoping to _ dynamically _ recover our 4d space time and in the second to test the factorization method against known analytical results . in all cases
we will be dealing with a system defined by a partition function @xmath3 and the corresponding phase quenched model @xmath4 where @xmath5 is the action of the system with its real and imaginary parts .
@xmath6 represents collectively the degrees of freedom of the model and in our case it corresponds to a set of @xmath7 matrices . in case
we are interested in measuring some observable @xmath8 , we consider the distribution functions @xmath9 and @xmath10 , where @xmath11 refers to @xmath12 .
then we define the fiducial system @xmath13 , the weight factor @xmath14 and the distribution @xmath15 factorizes @xmath16 where @xmath17
. then @xmath18 .
the @xmath19function constraint is implemented in our simulations by considering the system @xmath20 where @xmath21 and @xmath22 are parameters which control the constraining of the simulation .
the results are insensitive to the choice of @xmath23 as long as it is large enough .
then we have that @xmath24 .
the distribution of @xmath25 in @xmath26 has a peak @xmath27 and the quantity @xmath28 is the value of @xmath29 at @xmath30 .
the function @xmath31 can be obtained by integrating an analytic function to which we fit the @xmath32 data points . by applying this method
we force the system to sample configurations which give the essential contributions to @xmath33 , something that would be increasingly difficult with system size in the phase quenched model , eliminating this way the overlap problem .
this already allows us get close to the thermodynamic limit with modest computer resources .
furthermore we obtain direct knowledge of @xmath34 and @xmath15 which allows us to understand the effect of @xmath35 .
this is important for understanding the properties of the system when @xmath35 plays a crucial role . using the generic scaling properties of the weight factor @xmath36
, one may extrapolate the results obtained by direct monte carlo evaluations to larger system size .
such an extrapolation is expected to be particularly useful in cases where the distribution function turns out to be positive definite . in those cases we can
actually even _ avoid _ using the reweighting formula by reducing the question of obtaining the expectation value to that of finding the minimum of the free energy , which is ( minus ) the _ log _ of the distribution function . here
, the error in obtaining the scaling function propagates to the final result without significant magnifications .
therefore , the extrapolation can be a powerful tool to probe the thermodynamic limit from the accessible system size .
the so called ikkt matrix model @xcite has been conjectured to be a definition for non perturbative string theory .
a particularly interesting feature of the model is that the eigenvalues of the bosonic matrices @xmath37 generate space - time _
dynamically_. in our case we work with a reduced model where @xmath38 .
the observables that we study are the normalized eigenvalues @xmath39 of the space time `` moment of inertia '' @xmath40 .
an interesting scenario to investigate would be that the o(6 ) symmetry is spontaneously broken with some of the eigenvalues ( possibly 4 ) grow large and the rest remain small , providing a mechanism for dynamical compactification of extra dimensions .
* fig . 1 * : _ @xmath41 for n=128 , n = 16 . _
the ikkt model is a good example where finite size scaling for the oscillating factor @xmath36 works well ( the index @xmath42 corresponds to @xmath43 ) . for @xmath44
we find fast convergence to a scaling function @xmath45 , where @xmath46 @xcite .
we compute @xmath45 for small matrix size @xmath47 and the phase quenched distribution function @xmath48 for larger size @xmath49 .
then the factorization formula can be used in order to compute @xmath50 . note that in the computation of @xmath51 the errors do not propagate exponentially with system size since @xmath51 can be determined from the minimum of the `` free energy '' @xmath52 . in fig .
1 we show our results for @xmath53 for @xmath54 and @xmath55 .
a double peak structure of @xmath53 is evident and it is expected that the peak for small @xmath56 will increase with system size @xcite .
one hopes that this peak will be dominant for @xmath57 and that the large @xmath56 peak will be dominant for @xmath58 and this will realize the ssb scenario .
note the heavy suppression of the @xmath59 region caused by @xmath35 which is the peak of the phase quenched model distribution of @xmath60 .
+ * fig .
2 * : _ @xmath61 and @xmath62 for @xmath63 .
we consider rmt with one quark flavour and zero quark mass @xcite .
the model is chosen in order to study the correctness and effectiveness of the factorization method , since one can compare results with known analytical solutions even for finite @xmath49 .
the observable we measure is the `` quark number density '' @xmath64 as a function of the chemical potential @xmath65 , and we consider the distribution functions @xmath50 , where @xmath66 corresponds to the real and imaginary parts of @xmath64 respectively . notice that the effect of @xmath35 is dramatic , causing a discontinuous transition in @xmath64 . in @xcite
our results nicely reproduce the exact results known for finite @xmath49 and we are able to achieve large enough values of @xmath67 to obtain the thermodynamic limit . unfortunately , the function @xmath68 is not positive definite and the important contributions come from the region where it changes sign .
as expected , we find that finite size scaling does not work as well as in the case of the ikkt model ( although we obtain agreement up to order of magnitude for the values of @xmath69 that we explored ) .
we also find it very difficult to explore the crossover region near the phase transition point @xmath70 for @xmath71 since @xmath72 becomes very small .
since rmt is a schematic model of finite density qcd , we expect that the factorization method will be useful to explore the phase diagram of qcd
. 9 k. n. anagnostopoulos and j. nishimura , phys .
d * 66 * ( 2002 ) 106008 .
j. ambjrn et al . , jhep * 0210 * ( 2002 ) 062 .
v. azcoiti , g. di carlo , a. galante and v. laliena , phys .
* 89 * ( 2002 ) 141601 ; hep - lat/0210004 .
n. ishibashi , h. kawai , y. kitazawa and a. tsuchiya , nucl .
b * 498 * ( 1997 ) 467 ; h. aoki , s. iso , h. kawai , y. kitazawa and t. tada , prog .
* 99 * ( 1998 ) 713 .
e. v. shuryak and j. j. verbaarschot , nucl .
a * 560 * ( 1993 ) 306 ; j. j. verbaarschot and t. wettig , ann .
nucl . part .
* 50 * ( 2000 ) 343 . | we propose a method for monte carlo simulations of systems with a complex action .
the method has the advantages of being in principle applicable to any such system and provides a solution to the overlap problem . in some cases , like in the ikkt matrix model
, a finite size scaling extrapolation can provide results for systems whose size would make it prohibitive to simulate directly . |
this is the first in a sequence of papers in which we construct a quantum version of the map studied by kirwan @xcite , which maps the equivariant cohomology of a hamiltonian group action to the cohomology of the geometric invariant theory ( git ) quotient .
the existence of such a map suggested in the symplectic setting by salamon and ziltener @xcite , @xcite .
here we work under the assumption that the target is a smooth projectively - embedded variety with a connected reductive group action such that the stable locus is equal to the semistable locus ; this allows us to use the virtual fundamental cycle machinery of behrend - fantechi @xcite .
we prove that the quantum kirwan map intertwines the gromov - witten graph potential of the quotient with the _ gauged gromov - witten _ potential of the action in the large area limit . in physics language
, the quantum kirwan map relates correlators of a ( possibly non - linear , non - abelian ) gauged sigma model with those of the sigma model of the git quotient . as such
, the results overlap with those of givental @xcite , lian - liu - yau @xcite , iritani @xcite and others . the connection to mirror symmetry
is explained in the paper of hori - vafa @xcite : because mirror symmetry for vector spaces is rather trivial , the non - trivial change of coordinates arises when passing from a gauged linear sigma model to the sigma model for the quotient . since the quantum kirwan map is defined geometrically , it can be rather difficult to compute and the algebraic approach in @xcite , @xcite is more effective in cases where it applies .
however the geometric approach pursued here has the advantage over the approach in @xcite , @xcite that there are no semipositivity assumptions on @xmath4 or abelian - ness assumptions on the group @xmath2 .
also , @xmath1 can be a projective variety rather than a vector space .
to explain the gauged gromov - witten potential , recall that the _ equivariant cohomology _ of a @xmath2-space @xmath1 is the cohomology of the homotopy quotient @xmath5 where @xmath6 is a universal @xmath2-bundle .
equivariant quantum cohomology should count maps @xmath7 where @xmath8 is a curve equipped with some additional data and @xmath9 is a holomorphic map . any such map can be viewed as a map to @xmath10 together with a lift to @xmath11 .
holomorphic maps to @xmath12 correspond to holomorphic @xmath2-bundles , and so a holomorphic map to @xmath11 is given by a holomorphic @xmath2-bundle @xmath13 together with a holomorphic section of the associated @xmath1 bundle @xmath14 .
givental @xcite had earlier introduced an _ equivariant gromov - witten theory _ ,
base on equivariant counts of maps from a curve to @xmath1 .
these counts give rise to a family of products on the _ equivariant quantum cohomology _
@xmath15 where convergence issues are solved by the introduction of the _ novikov field _ @xmath16 . in the language of maps to the classifying space @xmath11 , givental s equivariant gromov - witten theory corresponds to counting maps from @xmath8 to @xmath11 whose image lies in a single fiber of the projection @xmath17 , that is , such that the @xmath2-bundle is trivial . in order to distinguish this theory from givental s
, we will call the theory with non - trivial bundles _ gauged gromov - witten theory _ , and call the map @xmath7 a _ gauged _ map to @xmath1 .
gauged gromov - witten invariants should be defined as integrals over moduli spaces of gauged maps . in order to obtain proper moduli spaces one needs to impose a _ stability _ or _ moment map _ condition as well as compactify the moduli space by , for example , allowing _
bubbling_. mundet and salamon s work , see @xcite , provides a symplectic approach to the moduli spaces of gauged maps and construction of invariants in the case that the target is a vector space .
thesis @xcite connects the symplectic approach to an algebraic stability condition , which combines the ramanathan stability condition for principal bundles with the hilbert - mumford stability for the action .
schmitt @xcite , @xcite had earlier constructed a grothendieck - style compactification in the case that the target is a smooth projective variety , while in the symplectic setting mundet @xcite and ott @xcite show that the connected components of the moduli spaces of semistable gauged maps have a kontsevich - style compactification .
in general , one needs virtual fundamental cycles to define integration over the moduli spaces .
here we restrict to the case that the target @xmath1 is a projective @xmath2-variety and note that a pair @xmath18 is by definition a morphism from @xmath8 to the _ quotient stack _ @xmath19 introduced by deligne - mumford @xcite .
we then use the theory of virtual fundamental classes developed by behrend - fantechi @xcite , based on earlier work of li - tian @xcite , to define gauged gromov - witten invariants .
the symplectic geometry is then only used as motivation , and to show that the deligne - mumford stacks that arise are proper .
of course it is desirable to have semistable reduction theorems to show properness but from the algebraic perspective even the stability conditions are somewhat obscure and we prefer the symplectic route to properness .
a good example is the moduli space of semistable bundles on a curve , where properness is immediate from the narasimhan - seshadri description as unitary representations of the fundamental group but semistable reduction is somewhat tricky .
the theory of gauged gromov - witten invariants in genus zero fits into an algebraic formalism that is a `` complexification '' of the theory of _ homotopy associative _ , or @xmath20 , algebras introduced by stasheff @xcite . in the @xmath20 story , which roughly corresponds to `` open strings '' in the mathematical physics language
, one has notions of @xmath20 algebras , @xmath20 morphisms , and @xmath20 traces .
these notions are associated with different polytopes called the associahedra , multiplihedra , and cyclohedra respectively .
the complexifications of these spaces are the grothendieck - knudsen space @xmath21 of stable @xmath22-marked genus @xmath23 curves , ziltener compactification @xmath24 of @xmath22-marked @xmath25-scaled affine lines , and the fulton - macpherson space @xmath26)$ ] of stable @xmath22-marked , genus @xmath23 maps of class @xmath27 $ ] to the projective line @xmath28 respectively .
the first space leads to the notion of _ genus zero cohomological field theory ( cohft ) _ , in particular , a _ cohft algebra _ given by the invariants with multiple incoming markings and a single outgoing marking .
the second space is associated to the notion of _ morphism of cohft algebras _ , and the third to the notion of _ trace on a cohft algebra_. the following is proved in gonzalez - woodward @xcite , and is reviewed in ( * ? ? ?
* theorem 7.20 ) .
[ gauged gromov - witten invariants ] @xcite [ theorem1 ] let @xmath1 be a smooth polarized projective @xmath2-variety and @xmath8 a smooth connected projective curve .
suppose that every semistable gauged map is stable ; then the category of stable gauged maps is a proper deligne - mumford stack equipped with a perfect obstruction theory .
the gauged gromov - witten invariants @xmath29 define a trace on @xmath0 .
the gauged gromov - witten invariants are also defined for polarized quasiprojective varieties under suitable properness assumptions for the moduli stacks of gauged maps , for example , if the polarization corresponds to an equivariant symplectic form with proper moment map convex at infinity . in particular ,
gauged gromov - witten invariants are defined for a vector space @xmath1 equipped with the action of a torus @xmath2 such that the weights are contained in an open half - space in the space of rational weights ; this condition implies in particular that the quotient @xmath4 is proper under the stable = semistable condition .
one can organize the gauged gromov - witten invariants into a formal _ gauged gromov - witten potential _
@xmath30 the splitting axiom implies that the bilinear form constructed from the second derivatives of the potential is compatible with the quantum product @xmath31 on @xmath32 in the usual sense .
however this bilinear form will usually be degenerate and so will not define a family of frobenius algebra structures .
it is convenient to throw into the definition of the trace certain _
liouville classes _ on the moduli space of gauged maps , in which case , if @xmath2 is trivial , @xmath33 becomes equivalent to the _ graph potential _ considered in givental @xcite . from the definition of mundet stability one
expects the gauged gromov - witten invariants to be related in the limit that the equivariant symplectic class @xmath34 $ ] approaches infinity to the gromov - witten invariants of the quotient @xmath4 , or more precisely , to the genus zero _ graph potential _ for the geometric invariant theory quotient @xmath4 @xmath35 for our purposes , it is more natural to work over the larger novikov field @xmath36 . from the symplectic point of view
the study of the large area limit of the vortex equations was initiated by gaio - salamon @xcite and ziltener @xcite , who suggested a map @xmath37 by counting `` affine vortices '' . by a hitchin - kobayashi correspondence with venugopalan @xcite ,
@xcite these maps correspond to the following algebraic objects : [ affinegauged ] ( affine gauged maps ) in the case that @xmath4 is a free quotient , an _ affine gauged map _ to @xmath1 consists of a tuple @xmath38 consisting of 1 .
( scaling form ) a meromorphic one - form @xmath39 on @xmath28 with only a double pole at @xmath40 ( hence inducing an affine structure on @xmath41 ) ; 2 .
( morphism to the quotient stack , stable at infinity ) a morphism @xmath42 , consisting of a @xmath2-bundle @xmath43 and a section @xmath44 such that @xmath45 is contained in the open subvariety @xmath46 associated with the semistable locus @xmath47 ; and 3 .
( markings ) an @xmath22-tuple of distinct points @xmath48 . in the case that the action of @xmath2 on the semistable locus of @xmath1 is only locally free we also allow a stacky structure at infinity .
that is , for @xmath49 an integer we denote by @xmath50 the group of @xmath51-th complex roots of unity and by @xmath52 : = ( { \mathbb{c}}^2 - \ { 0 \ } ) / { \mathbb{c}}^\times\ ] ] the weighted projective line with a @xmath50-singularity at @xmath53 , where @xmath54 acts on @xmath55 with weights @xmath56 .
the morphism @xmath9 is then required to be a representable morphism from some @xmath57 $ ] to @xmath19 .
[ quantum kirwan morphism ] [ large ] suppose that @xmath1 is a projective @xmath2-variety equipped with a polarization such that every semistable point is stable .
integrating over a compactified stack of affine gauged maps defines a morphism of cohft algebras @xmath58 by definition a morphism of cohft algebras consists of a sequence of maps @xmath59 satisfying a splitting axiom that guarantees that the formal map @xmath60 induces a @xmath61-homomorphism on each tangent space . in the case that the _ curvature _ @xmath62 of the quantum kirwan morphism vanishes , one obtains in particular a morphism of small quantum cohomology rings @xmath63 in this sense , morphisms of cohft algebras can be considered as non - linear generalizations of algebra homomorphisms .
more precisely the quantum kirwan morphism is defined by virtual integration over a compactification @xmath64 of the moduli stack of affine gauged maps equipped with evaluation and forgetful maps @xmath65 where @xmath66 is the rigidified inertia stack appearing in orbifold gromov - witten theory .
properness of this moduli space follows from compactness results of ziltener @xcite , @xcite ( for affine vortices ) and a hitchin - kobayashi correspondence for affine vortices due to venugopalan and the author , to appear .
this space again has a perfect relative obstruction theory , and pull - push using the virtual fundamental class gives rise to the maps @xmath67 . from the physics point of view
, witten @xcite explained the relationship between correlators in gauged sigma models and sigma models with target the symplectic quotient as a kind of `` renormalization '' given by `` counting pointlike instantons '' .
the `` quantum kirwan map '' is a precise mathematical meaning for this statement for arbitrary gauged ( possibly non - linear ) sigma models . despite the complicated - looking definition ,
the stack of affine gauged maps is easy to understand in simple cases .
( toric orbifolds ) let @xmath1 a vector space and @xmath2 a torus with lie algebra @xmath68 acting with weights @xmath69 contained in a half - space a real form @xmath70 .
assume that stable = semistable , so that the quotient @xmath4 is a proper toric deligne - mumford stack
. then @xmath71 is isomorphic to the stack of morphisms @xmath9 from @xmath72 to @xmath1 ( that is , @xmath1-valued polynomials in a single variable ) satisfying the following conditions for any @xmath73 : 1 .
( degree restriction ) the @xmath74-th component @xmath75 of @xmath9 has degree at most @xmath76 , that is , @xmath77 for some constants @xmath78 .
if @xmath79 is not an integer let @xmath80 .
2 . ( stability condition )
the collection of leading order coefficients @xmath81 lies in the semistable locus @xmath82 of @xmath1 and so defines a point in the substack of the inertia stack given by @xmath83 .
if @xmath4 is fano with minimal chern number at least two then @xmath84 and @xmath85 is an algebra homomorphism of small quantum cohomologies . thus knowing the map @xmath85 allows to give a presentation of the small quantum cohomology of @xmath86 . 1 . ( projective space )
if @xmath87 acts by scalar multiplication on @xmath88 then there is a unique morphism @xmath89 such that all components have degree @xmath25 , @xmath90 with marking at @xmath91 , given limit at infinity @xmath92 \in { \mathbb{p}}^{k-1 } = x { /\kern-.7ex/}g\ ] ] and vanishing value @xmath93 at @xmath94 .
this implies that @xmath95 where @xmath96 is the generator of @xmath0 and @xmath97 .
hence the presentation @xmath98/(\xi^k - q)\ ] ] of quantum cohomology of projective space .
( the teardrop orbifold , a weighted projective line ) if @xmath99 acts on @xmath100 with weights @xmath101 , so that @xmath102 $ ] is the teardrop orbifold , then @xmath103 implies that @xmath104 and @xmath105 is the point class . to evaluate @xmath106 interpret @xmath107 as the euler class of the vector bundles corresponding to the derivatives of @xmath9 at @xmath23 .
morphisms of class @xmath108 are given by pairs of polynomials @xmath109 integrating the euler class counts such maps whose derivatives @xmath110 zero at the marking and , if @xmath111 is even , semistable leading order terms @xmath112 \in { \mathbb{p}}[1,2 ] .\ ] ] are fixed to be a given point .
there is a unique such map .
for @xmath113 this implies @xmath114 half the generator @xmath115 of the twisted sector . for @xmath116
this implies that @xmath117 hence the presentation of the quantum cohomology of the teardrop orbifold @xmath118)= \lambda_x^g[\xi]/(q - 4\xi^3)\ ] ] which is a special case of coates - lee - corti - tseng @xcite .
see example 5.32 for more details .
a similar strategy gives relations for the quantum cohomology of any toric orbifold .
in addition , the setup also applies to quotients by non - abelian groups such as quiver varieties . in the large area limit
the graph potentials @xmath119 are naturally related via the quantum kirwan morphism : we fix a symplectic class @xmath34 \in h^2_g(x)$ ] and consider the stability conditions corresponding to the classes @xmath120 $ ] as @xmath121 varies . [ largearea ] suppose that @xmath8 is a smooth projective curve and @xmath1 a polarized projective @xmath2-variety such that stable = semistable for gauged maps from @xmath8 to @xmath1 of sufficiently large @xmath122 . then @xmath123 in other words , the diagram @xmath124{e , t}{{\kappa}_x^g } \node { } \node{qh ( x { /\kern-.7ex/}g ) } \arrow{sw , b}{\tau_{x { /\kern-.7ex/}g } } \\
\node { } \node{\lambda_x^g } \node { } \end{diagram}\ ] ] commutes in the limit @xmath125 .
the equality , or commutativity of the diagram , holds in the space of distributions in @xmath126 , in other words , for each power of @xmath126 separately .
mathematicians familiar with equivariant symplectic geometry in the last millennium will recognize the similarity with `` quantization commutes with reduction '' theorems , see for example guillemin - sternberg @xcite .
however , in the intervening years the use of `` quantum '' is mostly changed so now it refers to holomorphic curves .
the terminology `` adiabatic '' arises from the fact that stable gauged maps correspond to minima of an energy function depending on @xmath122 , so the theorem relates minima in the limit @xmath125 .
one is often interested not in the graph gromov - witten invariants but rather in the _ localized graph potential _ that is a solution @xmath127 to the quantum differential equation for @xmath4 : for @xmath128 @xmath129 , \quad \hbar \partial_\nu \tau_{x { /\kern-.7ex/}g,\pm } ( \alpha ) = \nu \star_\alpha \tau_{x { /\kern-.7ex/}g,\pm}(\alpha ) .\ ] ] the components of @xmath127 satisfy a version of the picard - fuchs equations which play an important role in mirror symmetry @xcite . the function @xmath130 resp .
@xmath131 can be expressed as the `` contribution from @xmath23 resp .
@xmath53 '' in the fixed point formula for the graph potential for @xmath28 induced by the circle action on @xmath28 .
there are similar gauged versions @xmath132\ ] ] which capture the contributions from @xmath133 to the localization formula applied to the _ gauged _ graph potential .
similarly the quantum kirwan map has an @xmath134-equivariant extension @xmath135 $ ] , with target formal power series with quotient modulo any power @xmath136 a polynomial in @xmath137 .
the factorization of the graph potentials generalizes to the gauged setting : @xmath138 and we prove a localized adiabatic limit theorem [ jresults ] below for the contributions to the fixed point formula .
let @xmath139 denote the classical kirwan map extended to @xmath3 by linearity over @xmath36 .
[ localized adiabatic limit theorem ] [ jresults ] [ jlargerel ] [ jlargearea ] in the setting of theorem [ largerel ] , @xmath140 in other words , after composition with the quantum kirwan map the localized graph potential is given by the classical kirwan map applied to the localized gauged potential .
the result also holds in the `` twisted case '' , that is , after inserting the euler class of the index of an equivariant bundle on the target , which in good cases describes the localized graph potentials of complete intersections . in this way one obtains formula similar to the `` mirror formulas '' of givental @xcite , lian - liu - yau @xcite , iritani @xcite and others for localized graph potentials @xcite in the toric case , but now for arbitrary geometric invariant theory quotients
however , the approach here is different from that of @xcite , @xcite etc . in that the `` mirror map '' is expressed as integrals over moduli spaces , while the approach of @xcite , @xcite etc .
solves for the `` mirror map '' as the solution to an algebraic equation .
the generality of these results allows various applications , mostly developed jointly with e. gonzalez , which will appear elsewhere .
we were rather surprised to discover that many of the `` standard formulas '' from classical equivariant symplectic geometry generalize to the quantum case by substituting the quantum kirwan morphism for the classical kirwan map ; this is rather unexpected since all of these formulas involve functoriality of cohomology in some way which is generally lacking in the quantum setting .
for example , we claim that there is ( i ) a wall - crossing `` quantum kalkman '' formula for gromov - witten invariants under variation of git , including invariance in the case of crepant flops ( ii ) an abelianization `` quantum martin '' formula relating gromov - witten invariants of quotients by connected reductive groups and their maximal tori , first suggested by hori - vafa ( * ? ? ? * appendix ) and bertram - ciocan - fontanine - kim @xcite and ( iii ) a quantum version of witten s non - abelian localization principle , relating the equivariant quantum cohomology correlators for @xmath1 with the quantum cohomology correlators for @xmath4 .
we remark that ciocan - fontanine - kim - maulik @xcite have introduced a different notion of gauged gromov - witten invariants of git quotients of affine varieties which works in any genus .
a symplectic interpretation of these invariants has recently been given by venugopalan @xcite .
from this point of view it seems that the ciocan - fontanine - kim - maulik invariants are a higher genus generalization of the `` large area chamber '' gauged gromov - witten invariants investigated here .
see @xcite , @xcite for more recent work .
to state the main result precisely , we have to explain what it means to have a `` commutative diagram '' of cohomological field theories . in this section
we describe the moduli spaces of stable curves ( complexified associahedron in genus zero ) , stable parametrized curves ( complexified cyclohedron ) and stable affine scaled curves ( complexified multiplihedron ) , which lead to the notion of cohft algebra , trace on a cohft algebra , and morphism of cohft algebras respectively , in analogy with the theory of @xmath20 spaces , morphisms , and traces .
then we introduce notions of compositions of morphisms and traces , or morphisms of cohft algebras , which are analogous to the composition of the corresponding @xmath20 notions .
this makes cohft algebras into a kind of @xmath53-category .
notably , we do not have a version of complexified multiplihedron for higher genus curves , which is why the theory here is restricted to genus zero .
moduli spaces of stable curves were introduced by mayer and mumford in 1964 @xcite , and further studied in deligne - mumford @xcite .
moduli of stable marked curves were studied by grothendieck in 1968 and later by knudsen @xcite . in this section
we describe these compactifications and the notion of _ cohomological field theory _ introduced by kontsevich - manin , see @xcite .
we remark that since the notion of stack is not introduced until ( * ? ? ?
* section 4 ) , we avoid it until then and adopt the point of view that the moduli spaces are just topological spaces .
first we describe stable curves .
[ stablecurves0 ] let @xmath141 be an integer
( nodal curves ) an _ @xmath22-marked nodal curve _ consists of a projective nodal curve @xmath8 with an @xmath22-tuple of distinct , smooth marked points @xmath142 .
an _ isomorphism _ of @xmath22-marked nodal curves @xmath143 is an isomorphism @xmath144 such that @xmath145 for @xmath146 .
2 . ( stable curves )
a nodal @xmath22-marked curve @xmath147 is _ stable _ iff @xmath8 has finite automorphism group .
that is , each genus zero component has at least three _ special points _ ( nodes or markings ) and each genus one component has at least one special point .
note that we do not require @xmath8 to be connected .
( modular graphs ) the combinatorial type @xmath148 of a stable curve is a _ modular graph _ : 1 . ( graph )
an unoriented graph @xmath149 whose vertices correspond to irreducible components of @xmath8 , finite edges @xmath150 to nodes , and semi - infinite edges @xmath151 to markings , equipped with a 2 .
( genus function ) @xmath152-valued function @xmath153 recording the genus of each irreducible component of @xmath8 and 3 .
( labelling of semi - infinite edges ) a bijection @xmath154 of the semi - infinite edges with labels @xmath155 . +
a modular graph is _ stable _ if it corresponds to a stable curve .
that is , each vertex with label @xmath23 resp .
@xmath25 has valence at least @xmath156 resp @xmath25 .
any stable curve @xmath8 has a _
universal deformation _ given by a family of curves @xmath157 over a parameter space
@xmath158 uniquely defined up to isomorphism , see for example @xcite .
let @xmath159 denote the set of isomorphism classes of connected genus @xmath160 , @xmath22-marked stable curves .
more naturally one should consider the moduli _ stack _ of stable curves but we put off discussion of stacks to ( * ? ? ?
* section 4 ) . in the case of genus
zero curves the universal deformation , and topology on @xmath159 , have a simple description @xcite , ( * ? ? ? * appendix d ) : for any marking @xmath161 and irreducible component @xmath162 of @xmath8 we denote by @xmath163 the node in @xmath162 connecting to the irreducible component of @xmath8 containing @xmath161 , or @xmath161 if @xmath161 is contained in @xmath162 .
( convergence of a sequence of stable curves ) a sequence @xmath164 $ ] converges to @xmath165 $ ] in @xmath159 if @xmath166 is isomorphic to @xmath167 for a sequence @xmath168 converging to @xmath169 in the base @xmath158 of the universal deformation .
explicitly , if @xmath170 , a sequence @xmath171 $ ] with smooth domain @xmath166 converges to @xmath172 $ ] if there exists , for each irreducible component @xmath162 of @xmath8 , a sequence of holomorphic isomorphisms @xmath173 such that 1 .
( limit of a marking ) for all @xmath174 , @xmath175 ; and 2 .
( limit of a different parametrization ) for all @xmath176 , @xmath177 has limit the constant map with value the node of @xmath162 connecting to @xmath178 . with the topology induced by this notion of convergence
, @xmath159 is compact and hausdorff ( and in fact , a projective variety @xcite . ) for any possible disconnected graph @xmath148 we denote by @xmath179 the space of isomorphism classes of curves of combinatorial type @xmath148 with @xmath22 semiinfinite edges and total genus @xmath160 , and @xmath180 its closure . if @xmath181 is a disjoint union then @xmath182 .
the moduli spaces of stable marked curves @xmath183 satisfy a natural functoriality with respect to morphisms of modular graphs @xmath148 .
( morphisms of modular graphs ) a _ morphism _ of modular graphs @xmath184 is a surjective morphism of the set of vertices @xmath185 obtained by combining the following : ( these are called _ extended isogenies _ in behrend - manin @xcite ) ( 0,0 ) # 1#2#3#4#5 @font ( 5874,1550)(2239,-2049 ) ( 2963,-1020)(0,0)[lb]@xmath186 ( 3828,-1020)(0,0)[lb]@xmath187 ( 7491,-1350)(0,0)[lb]@xmath188 1 .
@xmath189 _ collapses an edge _ if the map on vertices @xmath190 is a bijection except for a single vertex @xmath191 which has two pre - images connected by an edge in @xmath192 , and @xmath193 .
the genus function on @xmath194 is obtained by push - forward that is , @xmath195 .
+ ( 0,0 ) + # 1#2#3#4#5 @font + ( 3748,1128)(2239,-1627 ) ( 2767,-879)(0,0)[lb ] ( 5900,-879)(0,0)[lb ] 2 . @xmath189 _ collapses a loop _ if the map on vertices is a bijection and @xmath196 where @xmath197 is an edge connecting a vertex @xmath198 to itself .
then the genus functions on @xmath199 are identical except at @xmath198 where @xmath200 .
@xmath189 _ cuts an edge _ @xmath201 if the map on vertices is a bijection , but @xmath202 where @xmath203 are semiinfinite edges attached to the vertices contained in @xmath197 .
any morphism @xmath204 cutting an edge induces an isomorphism @xmath205 obtained by identifying the markings corresponding to the edges @xmath203 in @xmath194 .
+ ( 0,0 ) + # 1#2#3#4#5 @font + ( 5469,1280)(2239,-1779 ) ( 2837,-992)(0,0)[lb]@xmath160 ( 3612,-951)(0,0)[lb]@xmath23 ( 6900,-992)(0,0)[lb]@xmath160 4 .
@xmath189 _ forgets a tail _
( semiinfinite edge ) @xmath206 if the map on vertices is a bijection , but @xmath207 . in this case
there is a morphism @xmath208 obtained by forgetting the corresponding marking and collapsing any unstable components .
a morphism @xmath184 which collapses an edge or loop induces an inclusion @xmath209 of @xmath210 .
the morphisms arising in this way define on the collection of spaces @xmath211 the structure of a _ modular operad _
@xcite ; however , we will not need this language .
the boundary of @xmath179 in @xmath180 is the union of spaces @xmath212 such that @xmath148 is obtained from @xmath194 by collapsing edges and forgetting loops .
the boundary of @xmath159 is a union of the following subspaces ( which will be _ divisors _ with respect to the algebraic structure of the moduli space introduced later ) 1 .
( non - separating node ) if @xmath213 , a subspace @xmath214 equipped with an isomorphism @xmath215 the inclusion is obtained by identifying the last two marked points . 2 .
( separating node ) for each splitting @xmath216 with @xmath217 , a subspace @xmath218 corresponding to the formation of a separating node , splitting the surface into pieces of genus @xmath219 with markings @xmath220 , equipped with an isomorphism @xmath221 ( except that on the level of orbifolds in the case @xmath222 and @xmath223 there is an additional automorphism exchanging the components . )
the pull - back @xmath224 of any class @xmath225 has a knneth decomposition @xmath226 for some index set @xmath227 and classes @xmath228 .
each boundary divisor @xmath229 or @xmath230 has a homology class @xmath159 , and since @xmath159 is a rational homology manifold each of these homology classes has a dual class @xmath231 resp .
@xmath232 in @xmath233 . for the following ,
see manin @xcite .
[ cohft ] a _ cohomological field theory _ ( cohft ) with values in a field @xmath234 is a @xmath235-graded vector space @xmath236 equipped with a symmetric non - degenerate bilinear form @xmath237 and collection of @xmath238-invariant ( with koszul signs ) _ correlators _ @xmath239 where by convention @xmath240 if @xmath241 , satisfying the following two splitting axioms : 1 .
( non - separating node ) if @xmath242 then @xmath243 where @xmath244 are dual bases for @xmath236 ; 2 .
( separating node ) if @xmath245 , @xmath246 is a partition of @xmath247 , and @xmath248 with @xmath249 for @xmath250 then @xmath251 where the dots indicate insertion of the knneth components of @xmath252 , @xmath253 is a basis and @xmath254 a dual basis , and there is an additional factor of @xmath255 in the exceptional case @xmath223 , @xmath256 arising from the additional automorphism .
that is , if @xmath257 is as in then @xmath258 the correlators of any cohomological field theory define a family of associative algebra structures . in the standard axiomatization
these are part of the associated _ frobenius manifold structure _ on @xmath236 @xcite .
however for the purposes of functoriality it is helpful to keep the `` algebra '' and `` metric '' parts of this structure separate , and instead we define the following : a _ cohft algebra _
consists of a @xmath235-graded vector space @xmath236 and a collection of @xmath238-invariant ( with koszul signs ) multilinear maps @xmath259 satisfying the splitting axiom for any subset @xmath260 of order at least two : @xmath261 where @xmath262 is the knneth decomposition of the restriction of @xmath257 as in .
any cohft gives rise to a cohft algebra . by restricting to genus zero , and using duality to put one factor of @xmath236 on the right .
the splitting axiom is a special case of the splitting axiom in definition [ cohft ] . 1 .
manin @xcite terms a set of such maps a
_ @xmath263-structure _ on @xmath236 .
however we avoid this terminology since it is nt clear what this is a homotopy version of .
note that the cohft algebra structure does not require a metric , so a cohft may be thought of roughly speaking as a cohft algebra plus a metric .
we remark that a more natural definition of cohft may be obtained by allowing incoming and outgoing markings .
however , we prefer to write the classes on the left to emphasize the analogy with @xmath20 spaces .
3 . the various relations in @xmath264 give rise to relations on the maps @xmath265 .
in particular the map @xmath266 is associative , by the splitting axiom and the relation @xmath267 = [ d_{0 , \ { 0,1 \ } \cup \ { 2 , 3\ } } ] \in h^2({\overline}{m}_{0,4 } ) .\ ] ] 4 .
the notion of cohft algebra is the `` complex analog '' of the notion of @xmath20 algebra in the following imprecise sense .
denote by @xmath268 the moduli space of projective nodal curves @xmath8 equipped with an anti - holomorphic involution fixing the markings , that is , the moduli space of _ stable @xmath22-marked disks _ where the markings are not necessarily in cyclic order . the symmetric group @xmath238 acts canonically on @xmath268 by permuting the markings .
the quotient of @xmath268 by the action of @xmath238 is homeomorphic to the @xmath269-st associahedron @xmath270 introduced in stasheff @xcite .
@xmath20 space _ is a space @xmath1 equipped with a collection of maps @xmath271 satisfying a splitting axiom for the restriction to the boundary , and _
@xmath20 algebras _ arise as spaces of chains on @xmath20 spaces . to obtain the notion of cohft algebra we replace @xmath1 by a vector space and @xmath272 by the cohomology of its complexification .
one could also imagine a cochain - level version but there are reasons to expect that this gives nothing new , see teleman @xcite 5 .
the notion of a genus zero cohft can be repackaged in terms of a non - linear structure called a _ frobenius manifold _ , which is a non - linear generalization of the notion of _ frobenius algebra _
( unital algebra with compatible metric . )
any genus zero cohft ( with unit and grading ) gives rise to a frobenius manifold whose potential is @xmath273 the third derivatives of @xmath274 give rise to a family of algebra structures @xmath275 these give rise to a family of connections depending on a parameter @xmath137 , @xmath276 the associativity of @xmath61 translates into the flatness of @xmath277 , so that locally there exist sections @xmath278 satisfying @xmath279 for a full discussion of the correspondence between frobenius manifolds and cohft s the reader is referred to manin @xcite . in this section
we study the moduli spaces of stable marked _
parametrized _ curves .
these are a special case of moduli spaces of stable maps ( the degree one case ) but we prefer to view them in a different way , as a special case of the fulton - macpherson construction @xcite . we then discuss the associated notion of _ trace _ on a cohft algebra .
let @xmath8 be a smooth connected projective curve . 1 . ( parametrized nodal curves ) a _
@xmath8-parametrized nodal curve _ is a ( possibly disconnected ) nodal curve @xmath280 equipped with a morphism @xmath281 of homology class @xmath282 = [ c]$ ] and with the same arithmetic genus .
that is , @xmath280 is the union of irreducible components @xmath283 where @xmath9 maps the _ principal component _ @xmath284 isomorphically onto @xmath8 and @xmath9 maps the other irreducible _ bubble components _ @xmath285 onto points .
since the arithmetic genus of @xmath280 is the same as that of @xmath8 , the bubble components must be rational .
a _ marking _ of a @xmath8-parametrized curve is an @xmath22-tuple @xmath286 of points in @xmath287 distinct from the nodes and each other .
isomorphism _ of such curves is an isomorphism of nodal curves which induces the identity on @xmath8 .
( stable parametrized curves ) a @xmath8-parametrized curve is _ stable _ if it has no infinitesimal automorphisms , that is , each non - principal irreducible component of @xmath280 has at least three marked or nodal points .
( rooted forests ) any @xmath8-parametrized curve has a _ combinatorial type _ which is a forest @xmath148 ( finite collection of trees ) with a distinguished _ root vertex _ corresponding to the principal component and a labelling of the semiinfinite edges given by a bijection @xmath288 .
a rooted forest is _ stable _ if it corresponds to a stable parametrized curve , that is , each non - root vertex has valence at least three .
the set @xmath289 of isomorphism classes of connected stable @xmath8-parametrized curves has a natural topology , similar to that of @xmath290 in genus zero : the following can be taken as a definition or a proposition using a suitable construction of the universal deformation of a stable map to @xmath8 : ( convergence of a sequence of parametrized stable curves ) suppose @xmath8 has genus @xmath23 .
a sequence @xmath291 $ ] with smooth domain @xmath292 _ converges _ to @xmath293 $ ] if there exists , for each irreducible component @xmath294 of the limit @xmath280 , a sequence of holomorphic isomorphisms @xmath295 such that 1 .
( limit of a marking ) for all @xmath174 , @xmath296 , the node in @xmath162 connecting to the irreducible component of @xmath8 containing @xmath161 , or @xmath161 if @xmath161 is contained in @xmath162 ; 2 . ( limit of a different parametrization ) for all @xmath176 , @xmath177 has limit the constant map with value the node of @xmath162 connecting to @xmath178 ; and 3 . ( limit of the map ) for all @xmath74 , @xmath297
. convergence for nodal domains @xmath292 is defined similarly , by considering convergence on each irreducible component separately .
the definition for arbitrary genus is similar , but the maps @xmath298 exist only after removing small neighborhoods of the nodes .
the topology on @xmath289 induced by this notion of convergence is compact and hausdorff , and a special case of the fulton - macpherson compactification of configuration spaces considered in @xcite .
the open stratum @xmath299 of @xmath289 is the configuration space @xmath300 of @xmath22-tuples of distinct points on @xmath8 . more generally for any rooted forest @xmath148 with @xmath22 semiinfinite edges we denote by @xmath301 the moduli space of isomorphism classes with type @xmath148 and by @xmath302 its closure .
the moduli spaces of stable marked curves @xmath302 satisfy a natural functoriality with respect to morphisms of rooted forests @xmath148 .
we say that a _ morphism of rooted forests _ is a morphism of modular graphs corresponding to the rooted forests mapping the root vertex to the root vertex .
( morphisms of moduli spaces associated to morphisms of forests ) 1 .
any morphism of rooted forests @xmath184 induces a morphism of moduli spaces @xmath303 2 .
the boundary of @xmath301 is the union of spaces @xmath304 such that there is a morphism of rooted forests @xmath305 collapsing an edge .
if @xmath194 is obtained from @xmath148 by cutting an edge , then there is an isomorphism @xmath306 identifying the vertices corresponding to the additional semi - infinite edges .
the proof is standard from properties of moduli spaces of stable maps .
note that if @xmath307 is disconnected where @xmath308 contains the root vertex then @xmath309 where @xmath310 is the number of semi - infinite edges of @xmath311 .
thus the boundary of @xmath183 is the union of products of lower - dimensional moduli spaces of @xmath8-parametrized stable curves and stable curves .
( relation to the cyclohedron ) let @xmath8 be a projective line .
any anti - holomorphic involution of @xmath8 induces an anti - holomorphic involution of @xmath289 , with fixed point set @xmath312 identified with the moduli space of stable parametrized @xmath22-marked disks .
the symmetric group @xmath238 acts by permutation , and the quotient by @xmath313 is isomorphic to the subset @xmath314 of @xmath312 such that the marked points @xmath315 occur in cyclic order around the boundary of the disk .
the action of @xmath134 by rotation preserves @xmath314 and the quotient is the cyclohedron @xmath316 , that is , the moduli space of points on the circle compactified by allowing bubbling , see markl @xcite . in this sense
it is slight abuse of terminology to call @xmath289 the complexification of @xmath316 ; rather , @xmath289 is the complexification of a circle bundle over @xmath316 .
this dishonesty will be somewhat remedied in the last section when we consider @xmath289 with its circle action .
the boundary structure of the moduli space @xmath289 is described in the following .
the boundary of @xmath289 is the union of the following subspaces ( which will be divisors once the algebraic structure on @xmath289 is introduced ) : for each subset @xmath317 of order at least two a subspace @xmath318 where the markings for @xmath319 have bubbled off onto an ( unparametrized ) sphere bubble .
the subspace @xmath320 admits a homeomorphism @xmath321 for any @xmath322 , the pull - back @xmath323 to a subspace @xmath320 has a knneth decomposition @xmath324 for some index set @xmath227 and classes @xmath325 and @xmath326 . in general , the moduli space of stable maps is not smooth .
however , the space @xmath289 , as a special case of fulton - macpherson @xcite , is a compact smooth manifold . in particular
, any subset @xmath320 has a homology class @xmath327 \in h_2({\overline}{m}_n(c),{\mathbb{z}})$ ] and a dual class @xmath328 , although we work with rational coefficients below .
let @xmath234 be a vector space .
( trace on a cohft algebra ) [ cohfttrace ] a ( @xmath8-based , @xmath234-valued ) _ trace _ on a cohft algebra @xmath236 is a collection of @xmath238-invariant ( with koszul signs ) multilinear maps @xmath329 satisfying a splitting axiom @xmath330 where @xmath331 is the dual class to @xmath320 and the @xmath332 s denote insertion of the knneth components of @xmath257 .
that is , with @xmath257 as in , @xmath333 1 . in our main application ,
gauged gromov - witten invariants will define a @xmath36-valued trace on @xmath0 , which is a cohft defined over the field of fractions of @xmath334 . in other words , the space @xmath234 above need not be the ring or field over which the cohft algebra or cohft is defined .
2 . one should compare the notion of trace with that for @xmath20 algebras described in ( * ? ? ?
* proposition 2.14 ) . the corresponding notion for an @xmath20 space @xmath1
consists of a sequence of maps @xmath335 to an ordinary space @xmath336 , satisfying a suitable splitting axiom .
any trace @xmath337 on a cohft algebra @xmath236 defines a formal map @xmath338 often called a _ potential_. the splitting axiom implies that the second derivatives of @xmath339 with two point classes inserted define a @xmath234-valued family of bilinear forms on @xmath340 compatible with the multiplications @xmath341 on @xmath340 : ( family of bilinear forms associated to a cohft trace ) let @xmath342 denote the class given by the pull - back of a point under the @xmath74-th evaluation map @xmath343 . for @xmath344
define @xmath345 the _ bilinear form _ on @xmath340 associated to the trace .
mau - woodward @xcite , based on earlier work of ziltener @xcite , introduces a compactification of the moduli space of distinct points on the affine line up to translation , which `` complexifies '' the multiplihedron in the same way that the grothendieck - knudsen space and the fulton - macpherson spaces complexify the associahedron and cyclohedron respectively .
we then discuss the associated notion of _ morphism _ of cohft algebras .
let @xmath72 denote an affine line over @xmath346 , unique up to isomorphism .
we denote by @xmath347 the space of @xmath346-invariant one - forms on @xmath72 .
( scaled affine line ) [ scale ] a _ scaling _ of an affine line @xmath72 is a translation - invariant , non - zero one form @xmath348 .
a _ scaled affine line _ is an affine line equipped with a scaling .
an _ @xmath22-marking _ of an affine line is an @xmath22-tuple @xmath349 of distinct points in @xmath350 .
an _ isomorphism _ of scaled @xmath22-marked affine lines is an affine isomorphism @xmath351 , such that @xmath352 and @xmath353 .
( 0,0 ) # 1#2#3#4#5 @font ( 3441,1893)(2013,-3625 ) ( 3635,-3434)(0,0)[lb]@xmath354 ( 4840,-2624)(0,0)[lb]@xmath355 ( 4552,-1868)(0,0)[lb]@xmath356 ( 2518,-2571)(0,0)[lb]@xmath357 ( 2699,-1948)(0,0)[lb]@xmath358 ( 2946,-2243)(0,0)[lb]@xmath359 let @xmath360 denote the moduli space of isomorphism classes of scaled @xmath22-marked affine lines . if @xmath72 is a scaled affine line then the group of automorphisms of @xmath72 preserving the scaling is the additive group @xmath346 acting on @xmath72 by translation .
thus the moduli space @xmath360 may be identified with the configuration space @xmath361 of @xmath22-tuples of distinct points on @xmath72 up to the action of @xmath346 by translation , @xmath362 for any tuple @xmath363 we take the unique rescaling so that @xmath364 and then take the associated configuration .
conversely , any configuration defines an affine scaled map by taking the scaling @xmath364 to be standard .
[ twoform ] ( two - forms instead of one - forms ) the moduli space @xmath360 can be viewed in a different way : any scaling @xmath39 gives rise to a real area form @xmath365 on @xmath72 . replacing @xmath39 with @xmath366 amounts to forgetting a complex phase ; thus , one can view @xmath367 as the moduli space of data @xmath368 where @xmath369 are distinct points , @xmath370 is a translationally - invariant area form , and @xmath371 is a phase .
any automorphism @xmath372 of @xmath72 has a well - defined _ argument _ @xmath373 giving the angle of rotation , and @xmath372 acts on @xmath374 by @xmath375 .
this is the point of view taken in ziltener s thesis @xcite who suggested the motto `` one can not rotate affine symplectic vortices '' .
the moduli space @xmath360 has a natural compactification obtained by allowing bubbles with degenerate scalings .
( stable nodal scaled affine lines ) [ affinescaled ] a _ possibly degenerate scaling _ on an affine line @xmath72 is an element of the set @xmath376 a possibly degenerate scaling @xmath39 is _ degenerate _ if @xmath377 or @xmath378 and is _ non - degenerate _ otherwise .
the action of the group of automorphisms @xmath379 on @xmath380 by pull - back extends naturally to an action on @xmath381 , with fixed points @xmath382 .
let @xmath8 be a nodal curve and @xmath383 the dualizing sheaf , whose sections consist of one - forms possibly with poles at the nodes of @xmath8 whose residues on either side of a node are equal .
let @xmath384 denote the projectivization of @xmath383 .
let @xmath385 be a section .
a component @xmath386 of @xmath8 will be called _ colored _ if the the restriction of @xmath39 to @xmath386 is finite and non - zero .
nodal marked scaled affine line _ is a datum @xmath387 such that the following holds : 1 .
( monotonicity ) on any non - self - crossing path from a marking @xmath161 to the root marking @xmath354 , there is exactly one _ colored irreducible component _ with finite scaling ; and the irreducible components before ( resp .
after ) this irreducible component have infinite ( resp .
zero scaling ) .
see figure [ zilt ] , where irreducible components with infinite resp .
finite , non - zero resp .
zero scaling are shown with dark resp .
grey resp .
light grey shading .
an _ isomorphism _ of nodal marked scaled affine lines @xmath388 is an isomorphism of nodal curves @xmath389 intertwining the ( possibly degenerate ) scalings and markings in the sense that @xmath390 and @xmath391 .
a nodal marked scaled affine line is _ stable _ if it has no automorphisms , or equivalently , if each irreducible component with finite scaling has at least two special points , and each irreducible component with degenerate scaling has at least three special points .
the space @xmath392 of isomorphism classes of connected stable scaled @xmath22-marked lines has a natural topology , similar to the topology on the moduli space of stable curves .
given a stable affine scaled curve @xmath393 , for any marking @xmath161 and irreducible component @xmath162 we denote by @xmath163 the node in @xmath162 connecting to the irreducible component of @xmath8 containing @xmath161 , or @xmath161 if @xmath161 is contained in @xmath162 .
the following can be taken as a definition or a proposition with a suitable notion of family of stable scaled marked lines , see ( * ? ? ?
* example 4.2 ) .
( convergence of a sequence of nodal scaled affine lines ) a sequence @xmath394 $ ] with smooth domain @xmath166 _ converges _ to @xmath395 $ ] if there exists , for each irreducible component @xmath162 of the limit @xmath8 , a sequence of holomorphic isomorphisms @xmath173 such that 1 .
( limit of the scaling ) @xmath396 ; 2 . ( limit of a marking ) @xmath175 ; and 3 . ( limit of a different parametrization )
@xmath397 has limit the constant map with value the node of @xmath162 connecting to @xmath386 .
convergence for sequences with nodal domain is defined similarly .
( two markings converging ) if @xmath398 and two points @xmath399 come together in the sense that @xmath400 , then there exists a sequence of holomorphic maps @xmath401 such that @xmath402 converge to distinct points , and the scaling @xmath403 converges to zero .
the limiting configuration consists of a irreducible component with two markings and one node with zero scaling , and an irreducible component with finite scaling , one node , and the root marking @xmath354 .
see figure [ twoconverge ] .
( 0,0 ) # 1#2#3#4#5 @font ( 3569,1956)(3474,-4245 ) ( 6522,-3677)(0,0)[lb]@xmath354 ( 3955,-3660)(0,0)[lb]@xmath354 ( 3647,-3137)(0,0)[lb]@xmath355 ( 4172,-3148)(0,0)[lb]@xmath356 ( 6207,-2687)(0,0)[lb]@xmath355 ( 6553,-2699)(0,0)[lb]@xmath356 ( two markings diverging ) if @xmath404 for all @xmath405 with constant scaling @xmath406 and two points @xmath399 go to infinity in @xmath407 in different directions , then for @xmath408 there exists ( i ) a sequence of holomorphic maps @xmath409 such that @xmath410 and @xmath411 converge for @xmath412 and ( ii ) a sequence @xmath413 such that @xmath414 for @xmath415 converge to distinct points and @xmath416 converges to infinity .
the limiting configuration consists of two components with a single marking and node and finite scaling , and a component with two nodes , the root marking , and infinite scaling .
see figure [ twodiverge ] .
( 0,0 ) # 1#2#3#4#5 @font ( 4716,1697)(3474,-4292 ) ( 3955,-3660)(0,0)[lb]@xmath354 ( 3647,-3137)(0,0)[lb]@xmath355 ( 4172,-3148)(0,0)[lb]@xmath356 ( 7111,-3724)(0,0)[lb]@xmath354 ( 6366,-2992)(0,0)[lb]@xmath355 ( 7719,-2954)(0,0)[lb]@xmath356 we denote by @xmath392 the space of isomorphism classes of connected nodal scaled lines , equipped with the topology above . by mau - woodward @xcite @xmath392
is a compact hausdorff space .
( combinatorial types of nodal scaled affine lines ) the _ combinatorial type _ of a connected scaled affine line is a _ colored tree _ consisting of a tree @xmath417 together with a partition of the vertices @xmath418 and a labelling of its semi - infinite edges given by a bijection @xmath419 , and satisfying the combinatorial version of the monotonicity condition : 1 .
( monotonicity ) on any non - self - crossing path from a semi - infinite edge labelled @xmath74 to the semi - infinite edge labelled @xmath23 , there is exactly one vertex in @xmath420 , all vertices before resp .
after are in @xmath421 resp .
@xmath422 .
a colored tree is _
stable _ if it corresponds to a stable affine scaled line , that is , each vertex @xmath423 resp .
@xmath422 resp .
@xmath420 has valence at least @xmath156 resp . @xmath156
@xmath255 .
we call the vertices in @xmath420 the _ colored vertices_. colored trees can be pictured as trees where part of the tree containing the root semi - infinite edge edge is `` below water '' and part `` above water '' ; the colored vertices in @xmath420 are those `` at the water level '' .
the monotonicity condition then says that the path from any `` above water '' semiinfinite edge to the unique `` below water '' semiinfinite edge passes through the water surface exactly once .
however , in our trees we adopt the standard convention of having the root edge ( which corresponds to an outgoing marking ) at the top of the picture . more generally we allow disconnected curves where each connected component is either a nodal affine curve , or a nodal curve with infinite or zero scaling .
( colored forests ) a colored forests @xmath148 is a union of components that are either colored trees , or ordinary trees with all vertices in @xmath421 or all vertices in @xmath422 .
the semi - infinite edges of a colored forest admit a partition @xmath424 where @xmath425 is the union of semi - infinite edges labelled @xmath155 , and finite edges connecting @xmath421 with @xmath426 , and @xmath427 is the union of the semi - infinite edge with the set of edges connecting @xmath428 with @xmath422 .
[ edgepartition ] . for any colored forest @xmath148 with @xmath22 semiinfinite edges we denote by @xmath429 the space of isomorphism classes of scaled lines of combinatorial type @xmath148 , and @xmath430 its closure .
( morphisms of colored forests ) a _ morphism of colored forests _ from @xmath148 to @xmath194 is a combination of the following simple morphisms : 1 .
( collapsing edges without relations ) @xmath431 _ collapses an edge _ if @xmath189 is injective except that it maps two vertices in @xmath432 to the same vertex in @xmath433 or two vertices in @xmath422 with the same vertex in @xmath422 .
( in other words , any edge except those connecting @xmath420 with @xmath434 .
+ ( 0,0 ) + # 1#2#3#4#5 @font + ( 5921,1117)(2389,-866 ) 2 .
( collapsing edges with relations ) @xmath431 _ collapses edges _ if @xmath189 is injective except for one colored vertex of @xmath194 whose inverse image in @xmath435 is a collection of colored vertices in @xmath148 and a single vertex in @xmath422 , joined to each of the colored vertices by a single edge .
note that one can not write such a morphism as a composition of morphisms each collapsing a single edge , since there is no way to assign the coloring of vertices of the resulting graph which results in a colored forest .
+ ( 0,0 ) + # 1#2#3#4#5 @font + ( 5786,1092)(2389,-841 ) 3 .
( cutting an edge without relations ) @xmath431 _ cuts an edge _ of @xmath148 if the vertices are the same , but @xmath194 has one fewer edge than @xmath148 and the edge does not lie between the colored vertices and the root edge .
( cutting edges with relations ) @xmath431 _ cuts edges with relations _ of @xmath148 if the vertices are the same , but @xmath194 has fewer edges than @xmath148 , with each removed edge lying on a path between the root edge and the colored vertices , and each path passing through a unique such edge .
see figure [ cutwrelns ] .
( forgetting tails ) @xmath184 _ forgets a tail _
( semiinfinite edge ) and any vertices that become unstable , recursively starting from the semiinfinite edges furthest away from the root edge .
+ ( 0,0 ) + # 1#2#3#4#5 @font + ( 4916,1092)(2389,-803 ) ( more explanation on forgetting tails ) forgetting a tail leaves possibly only the vertex adjacent to the tail unstable , if it is colored with a single other edge adjacent , or non - colored with two other adjacent edges . in the first case , removing that vertex and the other adjacent edge still leaves a non - colored vertex which may be unstable , since it has one fewer edge .
if unstable , removing this vertex and identifying the other two edges gives a stable colored tree . in the second case
, removing the vertex gives a stable colored tree .
see figure [ forgetstages ] .
( morphisms of moduli spaces induced by morphisms of colored forests ) to any morphism @xmath189 of colored forests @xmath436 one can associate a morphism @xmath437 as follows . 1 . ( collapsing edges without relations ) any morphism @xmath184 collapsing an edge induces an inclusion @xmath438 .
2 . ( collapsing edges with relations ) any morphism @xmath439 collapsing edges with relations induces an inclusion @xmath438 .
3 . ( cutting an edge or edges with relations ) any morphism @xmath184 cutting an edge or edges with relations of @xmath148 induces a homeomorphism from @xmath430 to @xmath440 by identifying the markings corresponding to the additional semiinfinite edges .
( forgetting tails ) any morphism @xmath431 _ forgetting a tail _ induces a map @xmath441 which forgets the corresponding marking and collapses any unstable components recursively starting with the semiinfinite edges corresponding to the finite markings .
the existence of these maps is immediate from the definitions except for the existence of the morphism for forgetting tails , which requires an inductive argument collapsing the unstable components . for this note
that forgetting a tail leaves possibly only the component containing the corresponding marking unstable , if it had either non - degenerate scaling and two special points , or degenerate scaling and three special points . removing that component , and if the second possibility holds , replacing the node with the remaining marking or identifying the two remaining nodes , produces a new curve with one fewer irreducible component . in the second case ,
the resulting curve is automatically stable . in the first case
, the adjacent irreducible component has one fewer special point , and so now may be unstable .
if so , removing that component , and either ( i ) identifying the nodes , if the two special points were nodes , or ( ii ) changing the node to a marking , if the two special points were a node and a marking , produces a stable scaled affine curve .
we give a stronger , algebraic version of the forgetful morphism in ( * ? ? ?
* example 4.2 ) . for each @xmath148 ,
the boundary of @xmath430 consists of those moduli spaces @xmath442 such that @xmath148 is obtained from @xmath194 by collapsing an edge or edges with relations .
furthermore , each @xmath430 is a product of the moduli spaces @xmath443 and @xmath444 corresponding to the vertices of @xmath148 , where @xmath310 are the valences .
mau - woodward @xcite shows that the compactification @xmath392 has the structure of a projective variety , locally isomorphic to a toric variety .
the local structure of @xmath392 near the stratum @xmath445 of nodal lines with combinatorial type @xmath148 may be described as follows .
( balanced labellings ) for any colored tree @xmath148 , a labelling @xmath446 is _ balanced _ iff @xmath447 where @xmath448 range over elements of @xmath420 and @xmath449 is the unique non - self - crossing path from @xmath198 to @xmath450 , and in the product the sign is positive if @xmath197 is pointing towards the root edge marked @xmath354 and negative otherwise .
let @xmath451 denote the space of balanced labellings .
an element of @xmath452 is called a tuple of _ gluing parameters_. ( a singularity in the moduli space ) for the tree @xmath148 in figure [ cartierex2 ] , with large dots indicating vertices in @xmath420 , the relations are @xmath453 .
the corresponding toric variety @xmath452 corresponds to a @xmath156-dimensional cone with @xmath454 extremal rays , and so the moduli space has a singularity at the vertex . because toric surface singularities are always at worst orbifold singularities ,
the singularities of @xmath452 occur in complex codimension three and higher .
( 0,0 ) # 1#2#3#4#5 @font ( 4544,1864)(2389,-1613 ) ( 5277,-550)(0,0)[lb ] ( 2812,-1180)(0,0)[lb ] ( 3572,-1167)(0,0)[lb ] ( 5421,-1207)(0,0)[lb ] ( 6305,-1172)(0,0)[lb ] ( 3634,-550)(0,0)[lb ] @xcite [ maloc2 ] there exists an open neighborhood of @xmath445 in @xmath392 isomorphic to an open neighborhood of @xmath455 in @xmath456 the construction is a version of the construction of the universal deformation of a genus zero nodal curve @xcite in which small balls around the nodes are removed and the components glued together via maps @xmath457 .
the scaling is determined by the product of the gluing parameters from the root component to the irreducible components with finite scaling , independent of the choice of irreducible component with finite scaling by the balanced condition .
[ dimcount ] ( codimension formula ) the codimension of a stratum @xmath392 corresponding to a colored tree @xmath148 is _ not _ the number of finite edges , but rather @xmath458 where the extra summand @xmath459 corresponds to the minus the number of relations on the gluing parameters @xmath460 .
the boundary of @xmath392 consists of the following subsets ( which will be _ divisors _ with respect to the algebraic structure on @xmath392 introduced later ) : 1 .
( bubbling points ) for any @xmath461 of order at least two the subset @xmath462 corresponding to the formation of a single bubble containing the markings @xmath463 , with an isomorphism @xmath464 2 .
( blowing up scaling ) for any @xmath49 and unordered partition @xmath465 $ ] , @xmath466 of order at least two , with each @xmath467 non - empty , a subset @xmath468}$ ] corresponding to the formation of @xmath51 bubbles with markings @xmath469 , attached to a remaining component with infinite scaling .
we have a homeomorphism @xmath470 } \cong \left ( \prod_{i=1}^r { \overline}{m}_{|i_i|,1}({\mathbb{a } } ) \right ) \times { \overline}{m}_{0,r } .\ ] ] 1 .
the inclusions of these subspaces give the collection of spaces @xmath392 the structure of an algebra over the operad associated to the notion of _ homotopy morphism _ of operads in @xcite .
however , we will not use or need this language and will not discuss it further .
( relation to the multiplihedon ) the moduli space @xmath392 has a `` positive real locus '' that appears in stasheff s description of @xmath20 morphisms @xcite .
namely , taking a real structure on @xmath72 , the anti - holomorphic involution on @xmath72 induces an anti - holomorphic involution of @xmath392 .
we denote by @xmath471 the fixed point locus , in which all markings are on the real line .
the symmetric group @xmath238 acts on @xmath392 , and restricts to an action on @xmath471 with fundamental domain given as the closure of the subset @xmath472 where @xmath473 , homeomorphic to stasheff s multiplihedron @xmath474 @xcite .
an @xmath20 morphism of @xmath20 spaces @xmath475 consists of a sequence of maps @xmath476 satisfying a suitable splitting axiom on the boundary .
the splitting axiom for morphisms of cohft algebras is defined via divisors on @xmath392 . because the singularities of the toric variety @xmath452 occur in complex codimension at least three
, @xmath392 has a unique homology class of top dimension .
in particular , each of the boundary divisors above has a well - defined homology class in @xmath392 .
however , @xmath392 is not smooth ( and not a rational homology manifold ) and not every boundary stratum has a dual class .
that is , given a divisor @xmath477 } n_{[i_1,\ldots , i_r ] } d_{[i_1,\ldots , i_r ] } \ ] ] there may or may not exist a class @xmath478 that satisfies @xmath479 \rangle = \langle \beta \cup \gamma , [ { \overline}{{\mathcal{m}}}_{n,1}({\mathbb{a } } ) ] \rangle .\ ] ] this requires the restriction to combinations of boundary divisors that have dual classes in the following definition .
let @xmath480 and @xmath481 be cohft algebras . [ morphismcohfts ] a _ morphism of cohft algebras _ from @xmath236 to @xmath482 is a collection of @xmath238-invariant ( with koszul signs ) multilinear maps @xmath483 such that for any divisor @xmath484 of the form with dual class @xmath478 we have @xmath485 } n_{[i_1 , \ldots , i_r ] } \mu_w^s ( \phi^{i_1 } ( \alpha_i , i \in i_1 ; \cdot ) , \ldots , \phi^{i_r } ( \alpha_i , i \in i_r ; \cdot ) ; \cdot ) ( \iota_{[i_1,\ldots , i_r]}^ * \beta).\end{gathered}\ ] ] where the sum is over unordered partitions @xmath486 $ ] of @xmath487 with some @xmath467 possibly empty , @xmath332 indicates insertion of the knneth components of @xmath323 , @xmath488}^ * \beta$ ] , using the homeomorphisms , , the sum on the right - hand - side is assumed finite , and by convention if @xmath489 we replace @xmath490 with @xmath234 ( since in this case @xmath392 is empty ) .
a morphism of cohft algebras @xmath491 is _ flat _ resp .
_ curved _ if @xmath492 is zero resp .
non - zero .
[ drem ] 1 .
in our examples , @xmath493 will be a filtered ring , and @xmath494 filtered @xmath234-modules .
we say that a _ morphism of filtered cohft algebras _ is defined as above but where the right - hand - side of is finite modulo @xmath495 for any @xmath496 .
2 . see nguyen - woodward - ziltener @xcite for a description of the space of cartier divisors in @xmath392 , that is , a description of which combinations of codimension two strata have dual classes .
3 . the definition of flat morphism of cohft algebras ( which has nothing to do with flat morphism of rings etc . ) is analogous to the definition of flat morphism of @xmath20 algebras in @xcite .
that is , @xmath492 is analogous to the _ curvature _ of an @xmath20 morphism .
the divisors @xmath497 are points ( the limiting points in figures [ twoconverge ] , [ twodiverge ] ) and so have the same homology class . using the splitting axiom
this implies that if @xmath492 vanishes then @xmath498 is a homomorphism from @xmath499 to @xmath500 .
this is an analog of the fact that a flat @xmath20 morphism induces an algebra homomorphism of cohomology groups .
for simplicity we will consider here only the even case , that is , @xmath236 is a usual vector space and there are no signs .
now we discuss the connection of morphisms of cohft algebras with frobenius manifolds , or rather , the underlying family of algebras : let @xmath501 be vector spaces equipped with associative products @xmath502 varying smoothly in @xmath448 . a _ @xmath61-morphism _ from @xmath236 to @xmath482 is an analytic map @xmath503 whose derivative @xmath504 is a morphism of algebras from @xmath505 to @xmath506 for all @xmath507 .
in particular , if @xmath503 is a @xmath61-morphism with taylor coefficients @xmath508 then @xmath509 is an algebra homomorphism from @xmath510 to @xmath511 .
any morphism of cohft algebras @xmath512 from @xmath236 to @xmath482 defines a formal @xmath61-morphism from @xmath236 to @xmath482 via the formula @xmath513 @xmath491 arises from a flat morphism of cohft algebras iff @xmath514 . for convenience ,
we reproduce the argument from ( * ? ? ?
* proposition 2.43 ) .
consider the relation @xmath515 = [ d _ { \ { 1 \ } , \ { 2 \ } } ] \in h^2({\overline}{m}_{2,1}({\mathbb{a}}))\ ] ] from remark [ drem ] ( d ) .
its pull - back under the morphism @xmath516 forgetting all but the first two markings is the relation @xmath517 } [ d_{[i_1,i_2,\ldots , i_r ] } ] = \sum_{i } [ d_i]\ ] ] where the first sum is over partitions @xmath469 with @xmath518 , and the second is over subsets @xmath519 with @xmath520 . indeed , the pull - back of the coordinate on @xmath521 to @xmath24 is equal to any of the ( equal ) gluing parameters at the nodes at a generic point in any divisor @xmath522}$ ] appearing on the left - hand - side , see mau - woodward @xcite , and so has a zero of order one at that divisor . on the other hand ,
the pull - back is the inverse of the gluing parameter at a generic point in any divisor @xmath320 appearing on the right - hand side , and so has a pole of order one at that divisor .
the splitting axiom implies for each @xmath523 @xmath524 where the right - hand - side is assumed to be a finite sum in each graded piece . by definition
@xmath525 iff @xmath492 vanishes iff @xmath526 is flat . morphisms and traces on a cohft algebra admit a notion of composition , which generalizes the usual homotopy notions of composition in the @xmath20 setting .
the definition of @xmath255-morphism of a composition of a trace with a morphism depends on a moduli space of _ scaled parametrized curves _ which combines features of the complexified multiplihedron and cyclohedron .
let @xmath527 denote the space of @xmath22-marked @xmath25-scaled curves with underlying curve @xmath8 ; we do not quotient by automorphisms of @xmath8 .
the space @xmath527 admits a compactification @xmath528 by allowing _
stable scaled curves _ allowing bubbles with zero area form or allowing the area form on @xmath8 to degenerate to zero and _ affine scaled curves _ to develop as bubbles .
recall that any nodal map @xmath281 of class @xmath165 $ ] has a _ relative dualizing sheaf _ given as the tensor product of the dualizing sheaf @xmath529 for @xmath280 and the inverse of the pull - back of the cotangent bundle @xmath530 to @xmath8 : @xmath531 ( more detail is given below in ( * ? ? ?
* example 4.2 ) . )
a section of @xmath532 consists of a collection of rational one - forms @xmath533 on the bubble components , and a function @xmath534 on the principal component , such that @xmath535 have at most simple poles at the nodes and if so , the residues on either side of the node match .
we denote by @xmath536 the associated bundle with projective line fibers .
( nodal scaled marked curves ) [ scaledcurves ] let @xmath281 be a map of class @xmath165 $ ] .
scaling form _ is a section @xmath537 such that on any connected component @xmath538 of @xmath539 ( that is , bubble tree attached to the principal component ) the pair @xmath540 is an affine scaled curve ( if the scaling @xmath39 is infinite on @xmath284 ) or has zero scaling , otherwise .
a _ nodal scaled curve
_ parametrized by @xmath8 is a map @xmath541 equipped with a scaling form .
an _ isomorphism _ of nodal scaled parametrized curves @xmath542 , is an isomorphism @xmath543 such that 1 .
( scalings are intertwined ) @xmath544 .
( markings are intertwined ) @xmath545 ; 3 .
( parametrization is intertwined ) @xmath546 .
a nodal scaled parametrized curve is _ stable _ iff it has no infinitesimal automorphisms , that is , each irreducible non - principal component has at least three special points or a non - degenerate scaling and two special points .
the _ combinatorial type _ of a nodal scaled parametrized curve is a colored tree @xmath547 with finite resp .
semiinfinite edges @xmath548 resp .
@xmath549 obtained by replacing every irreducible component by a vertex and every node or marking with an edge , equipped with a _
root vertex _ @xmath550 corresponding to the principal component and a partition @xmath418 corresponding to the irreducible components with zero resp . finite resp
. infinite scalings .
a rooted colored tree is _
stable _ if it corresponds to a stable scaled curve , that is , every non - root vertex in @xmath421 or @xmath422 resp .
@xmath420 has at least @xmath156 resp
. @xmath255 incident edges . in other words ,
a stable scaled curve is a copy of the curve @xmath8 with a section of the trivial bundle ( necessarily constant ) and a collection of stable curves attached ( if the scaling on the principal component is zero or finite ) or a curve with infinite scaling and a collection of stable scaled affine lines attached ( if the scaling on the principal component is infinite ) .
let @xmath528 denote the moduli space of isomorphism classes of @xmath22-marked , scaled curves with principal component @xmath8 , and @xmath551 the subset of combinatorial type @xmath148 so that @xmath552 the topology on @xmath528 is similar to that for @xmath392 and is compact and hausdorff .
the local structure of @xmath528 near the stratum @xmath551 of nodal lines with combinatorial type @xmath148 may be described as follows .
as in , for any colored rooted tree @xmath148 , a labelling @xmath553 is _ balanced _ iff @xmath554 where @xmath448 range over elements of @xmath420 and @xmath449 is the unique non - self - crossing path from @xmath198 to @xmath450 , and the sign in the exponent is positive if @xmath197 is pointing towards the root vertex and negative otherwise .
let @xmath452 denote the set of balanced labellings .
as in @xcite for each rooted tree @xmath148 there exists a tubular neighborhood of the form @xmath555 given by removing small neighborhoods of the nodes and gluing together using identifications depending on the gluing parameters .
suppose that @xmath280 consists of a principal component @xmath556 with infinite scaling , two other components with infinite scaling ( dark shading ) , four components with finite scaling ( medium shading ) , and two components with zero scaling ( light shading ) as shown in figure [ grels ] .
the relations on the gluing parameters @xmath557 imply that the curve obtained with gluing with non - zero gluing parameters is equipped with the area form @xmath558 .
( 0,0 ) # 1#2#3#4#5 @font ( 4110,1553)(1899,-3608 ) ( 4840,-2624)(0,0)[lb]@xmath355 ( 4146,-2221)(0,0)[lb]@xmath356 ( 2763,-3161)(0,0)[lb]@xmath559 ( 3301,-2636)(0,0)[lb]@xmath560 ( 4400,-2483)(0,0)[lb]@xmath561 ( 4595,-3042)(0,0)[lb]@xmath562 ( 3374,-2949)(0,0)[lb]@xmath563 ( 4182,-3131)(0,0)[lb]@xmath564 ( 3086,-2416)(0,0)[lb]@xmath357 ( 5994,-2900)(0,0)[lb]@xmath565 , ( 5994,-3125)(0,0)[lb]@xmath566 ( 2084,-3076)(0,0)[lb]@xmath359 ( 3661,-3391)(0,0)[lb]@xmath284 ( 2572,-2723)(0,0)[lb]@xmath567 ( 2243,-3232)(0,0)[lb]@xmath568 ( 2015,-3208)(0,0)[lb]@xmath569 ( 2438,-2637)(0,0)[lb]@xmath358 ( 2430,-2504)(0,0)[lb]@xmath570 the boundary of @xmath528 is the union of the following sets ( which will be _ divisors _ once the algebraic structure on @xmath528 is introduced ) 1 .
( bubbling points ) for any subset @xmath571 of order at least two we have a subspace @xmath572 and an isomorphism @xmath573 corresponding to the formation of a bubble with markings @xmath574 with zero scaling .
( blowing up scaling ) for any unordered partition @xmath575 $ ] of @xmath576 we have a subspace @xmath577 } : d_{[i_1,\ldots , i_r ] } \to { \overline}{m}_{n,1}(c)\ ] ] and an isomorphism @xmath578 } : d_{[i_1,\ldots , i_r ] } \to { \overline}{m}_{r}(c ) \times \prod_{j=1}^r { \overline}{m}_{|i_j|,1}({\mathbb{a}})\ ] ] corresponding to degeneration of the area form to infinity , and bubbles with non - degenerate area form containing the markings .
( fixing a scaling ) for any @xmath579 , in particular for @xmath580 , there is an inclusion @xmath581 by choosing any scaling @xmath582 .
the description of boundary subspaces is immediate from the tubular neighborhood description of each stratum and a dimension count , which is the same as in remark [ dimcount ] .
the adiabatic limit theorem [ largearea ] will be deduced from the following divisor class relation : [ rhoequiv ] ( the basic divisor class relation in the moduli of stable scaled curves ) the homology class of @xmath583 is equal to that of the union of classes of @xmath468}$ ] over unordered partitions .
the equivalence in homology induced by the map @xmath584 equates the homology classes of @xmath585 with @xmath586 } d_{[i_1,\ldots , i_r]}$ ] , as one can check that the multiplicity of each divisor on the right hand side is @xmath25 , using the fact that linear equivalence of divisors implies homology equivalence .
suppose that @xmath587 are ( even , genus zero ) cohft algebras , with structure maps @xmath588 let @xmath589 be a morphism of cohft algebras , and @xmath590 traces on @xmath587 respectively .
[ 2morphism ] ( @xmath255-morphisms for compositions of traces and morphisms ) a _ @xmath255-morphism from @xmath591 to @xmath592 _ is a collection of maps @xmath593 such that 1 .
( fixing scaling ) if @xmath594 is the dual class to @xmath583 then @xmath595 2 .
( bubbling points ) if @xmath596 is the dual class to the divisor @xmath597 corresponding to bubbling off markings @xmath598 , then @xmath599 3 .
( blowing up scaling ) if @xmath600 } d_{[i_1,\ldots , i_r]}$ ] is a boundary divisor with dual class @xmath601 then @xmath602 } \tau^r_w ( \phi^{|i_1|}(\alpha_i , i \in i_1;\cdot),\ldots , \phi^{|i_r|}(\alpha_i , i \in i_r ; \cdot ) ; \cdot ) ( \iota_{[i_1,\ldots , i_r]}^ * \beta ) .\end{gathered}\ ] ] we write @xmath603 . if @xmath603 then @xmath604 } \tau^r_w ( \phi^{|i_1|}(\alpha_i , i \in i_1;\cdot),\ldots , \phi^{|i_r|}(\alpha_i , i \in i_r ; \cdot ) ; \cdot ) ( \iota_{[i_1,\ldots , i_r]}^ * \beta ) \ ] ] by combining ( a ) and ( c ) in the definition and using the equality of the homology class @xmath605 $ ] of the divisor @xmath606 corresponding to fixed scaling and the divisor at infinite scaling @xmath607}]$ ] .
if @xmath608 are the metrics on cohft algebras @xmath587 defined by traces @xmath590 and there is a @xmath255-morphism @xmath609 then @xmath610 is an isometry in the sense that @xmath611 we restrict to the case @xmath612 for simplicity . since @xmath603 we have @xmath613
in physics , a _ vortex _ refers to a stable solution of classical field equations which has finite energy in two spacial dimensions , see for example preskill @xcite and , for a more mathematical treatment , jaffe - taubes @xcite who classified vortices for scalar fields . in mathematics ,
vortices often refer to pairs of a connection and section of a line bundle satisfying an equation involving the curvature and a quadratic function of the section , see for example bradlow @xcite .
symplectic vortices are vortices in which the `` field '' takes values in a symplectic manifold with hamiltonian group action . in this section
we review the symplectic approach to gauged gromov - witten invariants , also known as symplectic vortex invariants or hamiltonian gauged gromov - witten invariants , as introduced by mundet and salamon , see @xcite , @xcite . if the moduli spaces of symplectic vortices are smooth , then integration over them defines the required invariants and the proofs of the theorems [ large ] , [ largearea ] are immediate
. of course , the moduli spaces are not smooth , or of expected dimension in general , and to define the needed virtual fundamental cycles we pass to algebraic geometry , starting in the following section . in this section
we review the construction of the moduli space of symplectic vortices as the symplectic quotient of the action of the group of gauge transformations on the space of gauged holomorphic maps ; this generalizes the construction of the space of flat connections as the symplectic quotient of the action of the group of gauge transformations on the space of connections on a bundle .
unfortunately since the space of gauged holomorphic maps is in general singular , this construction is rather formal and serves only for motivation for what follows .
let @xmath614 be a compact group with lie algebra @xmath615 .
we assume that @xmath615 is equipped with a @xmath614-invariant metric , inducing an identification @xmath616 .
we denote by @xmath617 a universal @xmath614-bundle , unique up to homotopy equivalence . for any @xmath614-space @xmath1 ,
we denote by @xmath618 the homotopy quotient and by @xmath619 the equivariant cohomology .
let @xmath13 be a principal @xmath614-bundle over a compact surface @xmath8 , and @xmath620 a classifying map for @xmath621 .
denote by @xmath622 the space of @xmath614-invariant forms and by @xmath623 the space of ( principal ) connections on @xmath621 , where @xmath624 is the generating vector field at @xmath625 .
the space @xmath626 is an affine space with a free transitive action of the space @xmath627 of one - forms with values in the _ adjoint bundle _
@xmath628 let @xmath629 denote the curvature map .
let @xmath1 be a hamiltonian @xmath614-manifold with symplectic form @xmath630 and moment map @xmath631 . by our convention
this means that for all @xmath632 , we have @xmath633 where @xmath634 is the generating vector field for @xmath96 .
recall that @xmath635 is the associated @xmath1-bundle .
sections @xmath636 are in one - to - one correspondence with lifts @xmath637 of @xmath372 to @xmath638 . given a section @xmath636 , the homology class @xmath639 $ ] is defined to be the homology class @xmath639 : = u_{k , * } [ c ] \in h_2^k(x,{\mathbb{z}}).$ ] the equivariant symplectic form @xmath640 pulls back to @xmath641 and descends to a closed , fiber - wise symplectic two - form @xmath642 depending on the choice of connection @xmath643 .
its cohomology class @xmath644 \in h^2(p(x))$ ] is independent of the choice of connection .
the _ equivariant symplectic area _ of @xmath9 is @xmath645 , [ \omega_k ] ) = ( [ c ] , u^ * [ \omega_{p(x ) } ] ) \ ] ] where @xmath646 denotes the pairing between homology and cohomology .
more concretely , we have @xmath647 independently of the connection @xmath643 .
we denote by @xmath648 the map induced by @xmath649 .
a _ gauged map _ from @xmath8 to @xmath1 is a datum @xmath650 where @xmath651 and @xmath636 is a section .
given a metric on @xmath8 , we denote by @xmath652 the associated hodge star .
the _ energy _ of a gauged map @xmath653 is given by @xmath654 suppose the @xmath8 is a complex curve .
denote by @xmath655 the space of almost complex structures on @xmath1 compatible with @xmath630 .
the action of @xmath614 induces an action on @xmath655 , and we denote by @xmath656 the invariant subspace .
any connection @xmath651 induces a map of spaces of almost complex structures @xmath657 by combining the almost complex structure on @xmath1 and @xmath8 using the splitting defined by the connection .
let @xmath658 denote the space of sections of @xmath635 .
we denote by @xmath659 the cauchy - riemann operator defined by @xmath660 .
( gauged holomorphic maps ) suppose that @xmath1 is a hamiltonian @xmath614-manifold equipped with an invariant almost complex structure @xmath227 , @xmath8 is a complex curve , and @xmath661 is a principal @xmath614-bundle .
gauged holomorphic map _ for @xmath621 is a pair @xmath653 satisfying @xmath662 .
let @xmath663 be the space of gauged holomorphic maps for @xmath621 : @xmath664 the energy and symplectic area are related by a generalization of the familiar energy - area relation for pseudoholomorphic maps in @xcite : ( energy - area relation , ( * ? ? ?
* proposition 3.1 ) ) suppose that @xmath1 is a hamiltonian @xmath614-manifold equipped with an invariant almost complex structure @xmath227 , @xmath8 is a compact complex curve , and @xmath13 is a principal @xmath614-bundle . let @xmath653 be a gauged map from @xmath8 to @xmath1 with bundle @xmath621 .
let @xmath383 be the area form determined by a choice of metric on @xmath8 .
the energy and equivariant symplectic area are related by @xmath665 the second term in this equation has a symplectic interpretation .
formally the space @xmath663 has a closed two - form induced from the sum of the symplectic form on the affine space of connections and the space of maps to @xmath1 .
let @xmath666 denote the fiber - wise two - form on @xmath635 defined by @xmath630 .
let @xmath667 be the area form defined by the metric and consider the formal two - form on @xmath663 @xmath668 where @xmath669 denotes the real - valued two - form induced by wedge product and the metric on @xmath670 .
the form is the formal restriction of a closed two - form on the space of all sections @xmath671 , and so restricts to a closed two - form on the smooth locus of @xmath663 . if @xmath1 is khler , then the moduli space of holomorphic sections , where smooth , is an almost complex manifold , and this can be used to show non - degeneracy of .
let @xmath672 denote the group of gauge transformations , with lie algebra the space of sections @xmath673 of the adjoint bundle @xmath670 with lie bracket defined by the map @xmath674 induced by lie bracket on @xmath615 .
the action of @xmath672 on @xmath663 has generating vector fields given by the covariant derivative and infinitesimal action @xmath675 where @xmath676 is induced by the infinitesimal action @xmath677 of @xmath614 .
the action preserves the two - form by invariance of the inner product @xmath646 on @xmath615 and symplectic form @xmath630 on @xmath1 . to prove the moment map condition , note that for @xmath679 and @xmath680 we have @xmath681 where @xmath682 resp .
@xmath683 denotes the derivative in the direction of @xmath198 resp .
@xmath684 .
[ vortexdef ] a gauged map @xmath685 is a _
symplectic vortex _ if it satisfies the @xmath686 an _ isomorphism _ of symplectic vortices @xmath687 with bundle @xmath621 is a gauge transformation @xmath688 such that @xmath689 and @xmath690 \circ u_0 = u_1 $ ] , where @xmath691 : p(x ) \to p(x)$ ] is the fiber - bundle - automorphism induced by @xmath491 .
@xmath22-marked _ symplectic vortex is a vortex @xmath653 together with an @xmath22-tuple @xmath692 of distinct points in @xmath8 .
a _ framed vortex _ is a collection @xmath693 , where @xmath694 is a marked vortex and @xmath695 is an @xmath22-tuple where each @xmath696 is a @xmath614-equivariant isomorphism , that is , a trivialization of the fiber .
let @xmath697 denote the moduli space of isomorphism classes of @xmath22-marked vortices and @xmath698 the union over isomorphism classes of bundles @xmath13 .
the moduli space @xmath699 is homeomorphic to the product @xmath700 where @xmath299 denotes the configuration space of @xmath22-tuples of distinct points in @xmath8 .
the bubbling phenomenon for holomorphic curves also occurs in the case of symplectic vortices and prevents compactness of the moduli spaces .
however , once one incorporates bubbles and fixes the homology class the moduli spaces become compact , as we now explain following salamon et al @xcite who proved compactness of the moduli space of vortices of fixed homology class in the case that @xmath1 has no holomorphic spheres , and mundet @xcite and ott @xcite who compactify the moduli space of vortices by allowing bubbling in the fibers of @xmath635 .
[ nodal ] ( nodal vortices ) suppose that @xmath1 is a hamiltonian @xmath614-manifold equipped with an invariant almost complex structure @xmath227 and @xmath8 is a connected smooth projective curve .
a _ nodal gauged @xmath22-marked map _ from @xmath8 to @xmath1 with underlying bundle @xmath621 consists of a datum @xmath701 where 1 .
( bundle and connection ) @xmath13 is a @xmath614-bundle and @xmath643 is a connection on @xmath621 ; 2 .
( stable section ) @xmath702 is a stable map of base degree one , that is , @xmath703 is a stable map of class @xmath165 $ ] ; 3 .
( markings ) an @xmath22-tuple @xmath704 of distinct , smooth points of @xmath280 .
a _ nodal vortex _ is a nodal gauged map such that the principal component is a vortex and there are no automorphisms with trivial bundle automorphism ( gauge transformation ) .
a nodal vortex is _ stable _ if it has finite automorphism group under the action of gauge transformations .
that is , a nodal gauged map is a vortex if the principal component is a vortex and each sphere bubble @xmath386 on which @xmath713 is constant has at least three marked or singular points .
note that there is no condition on the number of special points on the principal component .
in particular , constant gauged maps with no markings can be stable .
[ vortexconverge ] ( convergence of nodal vortices ) suppose that @xmath1 is a hamiltonian @xmath614-manifold equipped with an invariant almost complex structure @xmath227 and @xmath8 is a connected smooth projective curve .
suppose that @xmath718 is a sequence of marked nodal vortices on @xmath8 with values in @xmath1 , and @xmath701 is a nodal vortex with values in @xmath1 .
we say that @xmath719 $ ] _ converges _ to @xmath720 $ ] if after a sequence of bundle isomorphism @xmath721 the connections @xmath722 converge to @xmath20 in the @xmath723 topology and @xmath724 $ ] gromov converges to @xmath725 $ ] .
see ott @xcite for more details on the definition of convergence .
the definition of convergence implies in particular that the curvature @xmath726 converges to @xmath727 in @xmath728 , but @xmath722 is not required to ( and does not ) converge to @xmath643 uniformly in all derivatives .
recall from @xcite a condition which guarantees compactness of moduli spaces of symplectic vortices with non - compact target : a hamiltonian @xmath614-manifold @xmath1 with moment map @xmath729 is _ convex at infinity _ if there exists @xmath730 such that @xmath731 for every @xmath732 and @xmath733 outside of a compact subset of @xmath1 .
for example ,
if @xmath1 is a vector space with a linear hamiltonian action of @xmath614 with _ proper _ moment map @xmath649 then @xmath1 is convex .
the following is proved by mundet @xcite and ott @xcite .
[ compactnessforvortices ] ( sequential compactness for vortices with compact domain ) suppose that @xmath1 is hamiltonian @xmath614-manifold equipped with a proper moment map convex at infinity and an invariant almost complex structure @xmath227 , @xmath8 is a connected smooth compact complex curve .
any sequence of nodal symplectic vortices with bounded energy has a convergent subsequence . convergence for sequences with bounded first derivative is proved as follows : suppose that @xmath734 is a sequence of symplectic vortices on a bundle @xmath621 with smooth domain with the property that @xmath735 is bounded . by uhlenbeck compactness
, after gauge transformation , we may assume that @xmath722 converges weakly in the sobolev @xmath736 topology and strongly in @xmath723 to a limit @xmath20 .
after putting @xmath722 in coulomb gauge with respect to @xmath20 , elliptic regularity for vortices ( see @xcite ) implies that @xmath737 has a @xmath738-convergent subsequence of any @xmath739 .
more generally in the case of unbounded first derivative one has a bubbling analysis similar to that for pseudoholomorphic curves @xcite , @xcite : if @xmath740 and @xmath741 we say that @xmath742 is a _ singular point _ for the sequence @xmath734 . 1 .
( removal of singularities ) ( * ? ? ? * theorem 1.1 ) any finite energy vortex @xmath653 on the punctured disk @xmath743 extends to a vortex on @xmath484 .
( energy quantization ) ( * ? ? ?
* lemma 4.2 ) there exists an @xmath744 such that for any singular point of a sequence @xmath734 , one has @xmath745 .
( annulus lemma ) ( * ? ? ?
* section 5.2 ) there exists a constant @xmath746 such that if @xmath653 is a vortex on @xmath747 then for every @xmath748 there exist constants @xmath749 such that if @xmath750 and @xmath751 , then for @xmath752 the restriction of @xmath653 to @xmath753 satisfies @xmath754 4 .
( mean value inequality ) ( * ? ? ?
* corollary 2.2 ) let @xmath755 and let @xmath756 be an area form on the ball @xmath757 of radius @xmath51 around @xmath23 in @xmath346 . there exist constants
@xmath758 such that if @xmath653 is a vortex on @xmath757 with @xmath759 then @xmath760 we remark that an examination of the proof in @xcite shows that the constants @xmath761 depend only on bounds on @xmath39 and its first and second derivatives . given these ingredients
the proof of compactness goes as follows .
for each _ bubbling sequence _ @xmath762 with @xmath763 one constructs by _ soft rescaling _ a sequence of @xmath764-holomorphic maps @xmath765 on balls of increasing radius , with the property that the limit @xmath766 is zero and @xmath765 is an ordinary holomorphic map from @xmath346 to @xmath1 .
note that one does not have @xmath767 convergence of the sequence @xmath764 ; however , ( * ? ? ?
* appendix ) shows that @xmath723 convergence of a sequence of almost complex structures is sufficient as long as one has a version of the mean value inequality , which in this case follows from the vortex equations .
the limiting configuration is then constructed by induction .
denote by @xmath768 the set of isomorphism classes of nodal vortices with connected domain .
the _ combinatorial type _ of a nodal vortex is a rooted tree @xmath148 with root vertex corresponding to the principal component , equipped with a labelling of vertices by elements of @xmath769 . for any tree @xmath148 and homology class @xmath770
we denote by @xmath771 the space of isomorphism classes of vortices of combinatorial type @xmath148 , so that @xmath772 as @xmath148 ranges over connected combinatorial types .
we say that a subset @xmath158 of @xmath768 is _ closed _ if any convergent sequence in @xmath158 has limit point in @xmath158 , and _
open _ if its complement is closed .
the open sets form a topology for which any convergent sequence is convergent , and any convergent sequence has a unique limit , by arguments similar to ( * ? ? ?
* lemma 5.6.5 ) .
namely local `` distance functions '' can be defined by combining the distance functions of @xcite with the @xmath728-metric on the space of connections . in the case
@xmath489 , given a constant @xmath746 and stable vortex @xmath773 and another stable vortex @xmath774 with the same underlying bundle @xmath621 define @xmath775,[(p , a_1,\hat{c}_1,{u}_1 ) ] ) = \inf_{k \in { \mathcal{k}}(p ) } \vert k \cdot a_1 - a_0 \vert_{l^2 } + { { \operatorname}{dist}}_{\epsilon}^0 ( [ ( \hat{c}_0,u_0 ) ] , [ ( \hat{c}_1,k \cdot u_1)])\ ] ] where @xmath776 is the distance on isomorphism classes of stable maps defined on @xcite , but using the yang - mills - higgs energy on a small ball around each node .
it follows from the results in ott @xcite and elliptic regularity for vortices that for @xmath777 sufficiently small , a sequence @xmath778 $ ] converges to @xmath779 $ ] if and only if @xmath780,[(p , a_\nu,\hat{c}_\nu,{u}_\nu ) ] ) \to 0\ ] ] and the remainder of the proof is similar to that in @xcite . the sequential compactness theorem [ compactnessforvortices ] implies : ( properness of the moduli space of symplectic vortices ) suppose that @xmath1 is a hamiltonian @xmath614-manifold equipped with a proper moment map convex at infinity and an invariant almost complex structure @xmath227 and @xmath8 is a connected smooth projective curve .
then @xmath768 is hausdorff and the energy map @xmath781 is proper . in the large area limit
the vortices are related to holomorphic maps to the ( possibly orbifold ) symplectic quotient , as pointed out by gaio - salamon @xcite .
the limiting process involves various kinds of bubbling which one hopes to incorporate into a description of the relationship .
first recall the notion of symplectic quotient of @xmath1 by @xmath614 as introduced by mayer and marsden - weinstein : suppose that @xmath1 is a hamiltonian @xmath614-manifold equipped with a proper moment map @xmath782 .
let @xmath783 denote the symplectic quotient .
if @xmath614 acts freely and properly on @xmath784 , then @xmath785 has the structure of a smooth manifold of dimension @xmath786 with a unique symplectic form @xmath787 satisfying @xmath788 , where @xmath789 and @xmath790 are the inclusion and projection respectively .
any invariant almost complex structure @xmath227 on @xmath1 induces an almost complex structure on @xmath785 .
if @xmath791 is a smooth projectively embedded variety in a @xmath2-representation @xmath236 then @xmath785 is canonically homeomorphic to the geometric invariant theory quotient @xmath4 introduced by mumford , by a theorem of kempf - ness .
the latter is defined as the quotient of the _ semistable locus _
@xmath792 where @xmath793 is the @xmath794-th tensor product of the hyperplane bundle on @xmath1 .
the git quotient @xmath4 is the quotient of @xmath82 by the _ orbit - equivalence relation _ @xmath795 a point @xmath732 is _ stable _ if @xmath796 is closed in @xmath82 and the stabilizer @xmath797 is finite .
if stable = semistable in @xmath1 then the points of @xmath4 are the orbits of @xmath2 in @xmath82 , that is , two orbits are equivalent iff they are equal .
see gaio - salamon ( * ? ? ?
* section 2 ) . given an infinite - area vortex @xmath653 , let @xmath804 denote the composition of @xmath9 with the quotient map @xmath805 .
the equation @xmath806 implies @xmath807 ( since @xmath808 is holomorphic ) and so @xmath809 .
conversely , given @xmath804 let @xmath621 be the pull - back of @xmath810 , equipped with the connection given by @xmath811 .
the equivariant map @xmath812 defines a section @xmath813 .
the equation @xmath807 implies @xmath814 , since @xmath660 agrees with @xmath815 on @xmath816 .
if @xmath653 is constructed in this way from @xmath817 then the corresponding map to @xmath785 is @xmath817 . to see that
the map @xmath818 is injective , suppose that @xmath653 and @xmath819 define the same holomorphic map to @xmath785 . then after gauge transformation @xmath820 and the equations @xmath821 and local freeness of the action imply that @xmath822 .
( sequential compactness for bounded first derivative ) suppose that @xmath1 is a hamiltonian @xmath614-manifold with proper moment map convex at infinity equipped with an invariant almost complex structure @xmath227 and @xmath8 is a connected smooth projective curve .
suppose that @xmath734 is a sequence of symplectic vortices for @xmath823 with the property that @xmath825 is bounded .
then there exists a subsequence and a sequence of gauge transformations @xmath826 such that @xmath827 converges to an infinite - area vortex @xmath828 uniformly in all derivatives .
( bubble zoology for the infinite area limit ) suppose that @xmath1 is a hamiltonian @xmath614-manifold with a proper moment map convex at infinity and an invariant almost complex structure @xmath227 and @xmath8 is a connected smooth projective curve .
suppose that @xmath734 is a sequence of vortices for @xmath829 with @xmath830 consider the _ rescaled pair _
@xmath831 on @xmath832 where @xmath833 .
noting that @xmath834 has @xmath835 bounded .
then after passing to a subsequence one of the following possibilities occurs : 1 .
( sphere bubble in @xmath785 ) if @xmath836 , then @xmath834 converges to a solution to a solution to [ limeq ] , that is , is equivalent to a holomorphic map to @xmath837 .
( affine vortex ) if @xmath838 , then @xmath834 converges to a vortex on the affine line @xmath72 , with respect to the euclidean area form @xmath839 .
3 . ( sphere bubble in @xmath1 ) if @xmath840 , then @xmath834 converges to a vortex with zero vortex parameter , that is , a holomorphic map @xmath841 . in this section
we further study vortices on the complex affine line @xmath72 , which arose in the gaio - salamon study @xcite of the large area limit .
ziltener @xcite studied the bubbles that arise in more detail .
( affine vortices ) suppose that @xmath1 is a compact hamiltonian @xmath614-manifold equipped with an invariant almost complex structure @xmath227 .
@xmath22-marked affine symplectic vortex _ with target
@xmath1 is a datum @xmath694 , where @xmath842 is a connection on the trivial bundle @xmath843 , @xmath89 is a @xmath660-holomorphic map , @xmath844 is an @xmath22-tuple of distinct points , and @xmath845 an _ isomorphism _ of scaled vortices @xmath846 to @xmath847 is an automorphism @xmath491 of @xmath848 ( necessarily a translation ) and a gauge transformation @xmath849 satisfying @xmath850 .
let @xmath851 denote the moduli space of isomorphism classes of @xmath22-marked affine vortices .
it has a natural compactification by isomorphism classes of _ nodal affine vortices _ described as follows .
let @xmath8 be an @xmath852-marked connected genus zero nodal curve with irreducible components @xmath853 .
for each irreducible component @xmath386 not containing @xmath354 there is a unique node @xmath854 which disconnects @xmath386 from the marking @xmath354 ; we denote by @xmath855 the complement . in case @xmath386
contains @xmath354 we denote @xmath856 .
each affine curve @xmath857 is isomorphic to @xmath346 , uniquely up to translation and dilation .
thus @xmath857 admits a unique equivalence class of khler forms , equal to @xmath858 up to scalar multiplication .
( nodal affine vortices ) suppose that @xmath1 is a compact hamiltonian @xmath614-manifold equipped with an invariant almost complex structure @xmath227 .
@xmath22-marked nodal affine symplectic vortex _ with target
@xmath1 is a datum @xmath859 consisting of a connected @xmath852-marked nodal curve @xmath8 together with a principal @xmath614-bundle @xmath13 , a ( possibly infinite or zero ) two form @xmath860 , a connection @xmath861 on each @xmath862 , a section @xmath863 , and markings @xmath864 , such that each irreducible component @xmath386 of @xmath8 is one of the following three types : 1 .
( zero scaling ) components with zero two - form , equipped with a trivial bundle @xmath865 and a pseudoholomorphic map @xmath866 .
( finite , non - zero scaling ) components with non - zero , finite area form @xmath867 equal to a non - zero multiple of @xmath366 on @xmath868 and an affine vortex @xmath869 on @xmath857 .
3 . ( infinite scaling ) components @xmath386 with infinite two - form @xmath870 , equipped with holomorphic sections @xmath871 mapping to the zero level set @xmath872 , and so defining a holomorphic map @xmath873 1 . ( monotonicity ) for every non - self - crossing path from a marking @xmath874 to a marking @xmath354 , the path crosses exactly one irreducible component with finite , non - zero area form , and all irreducible components before resp .
after that irreducible component have zero resp . infinite area form
. 2 . ( continuity )
if @xmath386 meets @xmath162 at a node represented by a pair @xmath875 then @xmath876 .
3 . ( stability ) if @xmath386 is an irreducible component on which the two - form is zero or infinity resp .
finite and non - zero and @xmath713 is constant resp .
@xmath713 is covariant constant and @xmath861 is flat then @xmath386 contains at least two resp .
three special points .
an irreducible component @xmath386 is a _ ghost component _ if it satisfies one of the hypotheses requiring at least three special points or non - degenerate scalings ; that is , @xmath877 is constant or @xmath878 is flat and @xmath879 is covariant constant .
the stability condition can then be reformulated as the condition that any ghost component has at least three special points ( nodes or markings ) or two special points and a non - zero , finite area form .
either of these conditions is equivalent to the absence of non - trivial infinitesimal automorphisms : infinitesimal automorphisms arising from gauge transformations are impossible because of the local freeness assumption for the action of @xmath614 on @xmath802 .
( convergence of nodal affine vortices ) a sequence of isomorphism classes of nodal marked affine symplectic vortices @xmath880 $ ] _ converges _ to a nodal marked affine vortex @xmath881 $ ] iff there exists for each irreducible component @xmath162 of @xmath8 , a sequence @xmath882 of increasing open neighborhoods and for each @xmath405 , a holomorphic embedding @xmath883 , an isomorphism @xmath884 such that if @xmath885 denotes the associated maps of fiber bundles then 1 .
( open neighborhoods cover ) the union of @xmath886 is @xmath166 ; 2 .
( marked curves converge ) if @xmath887 then the limit of @xmath888 is defined and equal to @xmath161 . furthermore , if @xmath889 then the image of @xmath890 converges to the node of @xmath386 connecting to @xmath162 ; 3 .
( connections converge ) @xmath891 converges to @xmath892 uniformly in all derivatives in any compact subset of any @xmath893 ; 4 .
( sections converge ) @xmath894 converges to @xmath895 uniformly in all derivatives in any compact subset of any @xmath896 ; 5 .
( scalings converge ) @xmath897 converges to @xmath898 in @xmath723 on any compact subset of any @xmath893 ; 6 .
( energies converge ) @xmath899 ( sequential compactness for affine vortices ) [ affinecompact ] suppose that @xmath1 is a hamiltonian @xmath614-manifold with a proper moment map convex at infinity and an invariant almost complex structure @xmath227 .
any sequence of isomorphism of classes of stable marked affine vortices with bounded energy has a convergent subsequence .
( bubbling analysis for affine vortices ) [ affinevortices ] [ annlem2 ] suppose that @xmath1 is a hamiltonian @xmath614-manifold with proper moment map convex at infinity equipped with an invariant almost complex structure @xmath227 . 1 .
( energy quantization ) ( * ? ? ?
* lemma d.1 ) there exists a constant @xmath900 such that any non - trivial symplectic vortex @xmath901 on @xmath346 satisfies @xmath902 .
[ al ] ( annulus lemma ) ( * ? ? ?
* lemma 4.11 ) for every compact subset @xmath903 and every number @xmath904 there are constants @xmath905 and @xmath906 such that the following holds .
assume that @xmath907 and @xmath653 is a vortex on the annulus @xmath908 such that @xmath909 for every @xmath910 , and suppose that @xmath911
. then for every @xmath912 we have @xmath913 3 . [ mvt ]
( mean value inequality ) let @xmath903 be a compact subset .
then there exists a constant @xmath900 such that for every @xmath914 , @xmath755 and every symplectic vortex @xmath653 satisfying @xmath915 and @xmath916 , the energy density @xmath917 given by the integrand in satisfies the estimate @xmath918 4 .
( removal of singularities ) ( * ? ? ?
* proposition d.6 ) [ k0 ] let @xmath653 be a finite energy vortex on @xmath346 .
the map @xmath919 extends continuously to a map @xmath920 , such that @xmath921 .
furthermore , 1 .
there are constants @xmath922 and @xmath923 such that the following holds . for every vortex @xmath653 on @xmath346 and every @xmath924 such that @xmath925 and every @xmath926 we have @xmath927
there exist a number @xmath923 such that for @xmath928 , then @xmath929 exists , and there exists a map @xmath930,g)$ ] such that if @xmath931 denotes the restriction of the connection @xmath643 in radial gauge to the circle @xmath932 , then @xmath933 @xmath934 necessarily @xmath935 is fixed by @xmath936 , which since @xmath614 is compact and acts locally freely on @xmath802 , is finite order
. [ extend ] the compactness argument for some of the moduli spaces below uses a slightly more general argument than that stated in ziltener @xcite and ott @xcite .
namely , multiples of the standard area form on the projective line gives rise to area forms in local coordinates @xmath937 for @xmath746 .
we will need the ( annulus lemma ) above and the associated exponential decay of the derivative of a vortex for this area form . to see that the result still holds , we remark that ziltener s results on the invariant symplectic action in @xcite hold for any area form @xmath938 satisfying the following property : let @xmath939 denote the minimal length of generating vector fields for lie algebra vectors of length one .
we say that @xmath940 is _ admissible _ for some constant @xmath941 iff @xmath942 see ziltener ( * ? ? ?
* equation 1.7 ) , in particular , for the area form . the mean value inequality in ott ( * ? ? ?
* section 2 ) holds uniformly in @xmath777 for the family of area forms .
thus the inequality in the ( annulus lemma ) implies an inequality for the covariant derivative of the map away from the ends . in the case that @xmath614 acts freely on @xmath802 , any finite energy vortex @xmath653 on @xmath346 has a well - defined _ evaluation at infinity _
@xmath943 , given by the limit of @xmath944 along any ray @xmath945 . more generally in the locally free case ,
@xmath946 $ ] lies in the _ inertia orbifold _ @xmath947 which , for example , appeared in kawasaki @xcite .
the orbifold case was not treated in ziltener @xcite ; however , the proof is almost the same as the manifold case .
[ extends ] ( failure of removal of singularities ) the area form @xmath366 has a pole of order two at infinity and so one can not expect an extension of @xmath653 to the projective line to satisfy the vortex equations .
instead , the last item implies that the connection @xmath643 converges in @xmath948 , after gauge transformation , to a connection of the form @xmath949 near infinity for some @xmath950 with @xmath951 finite order . by the implicit function theorem on @xmath952 using weighted sobolev spaces
, one sees that any connection with exponential decay to @xmath949 is complex gauge equivalent to @xmath949 , for @xmath953 sufficiently large .
thus , after _ complex _ gauge transformation , we may take @xmath643 to be equal to @xmath949 in a neighborhood of @xmath40 . taking @xmath621 to be a orbi - bundle on @xmath954 given by gluing in the trivial bundle using transition map @xmath955
, it follows that @xmath653 extends to a gauged holomorphic map on @xmath28 mapping @xmath53 to @xmath802 .
however , the extension will not satisfy any vortex equation on the entire projective line @xmath28 .
let @xmath956 resp .
@xmath957 denote the moduli space of isomorphism classes of nodal scaled affine vortices to @xmath1 , resp .
the moduli space of isomorphism classes of framed nodal scaled affine vortices to @xmath1 .
@xmath958 admits an evaluation maps at the markings , and , as explained in @xcite an additional evaluation map at infinity to @xmath959 : @xmath960 if @xmath961 acts freely , combining this map with a classifying map gives a map @xmath962 if @xmath963 only acts locally freely , then the map above exists as a morphism of stacks , or after passing to a classifying space for the groupoid @xmath956 , which has the same rational cohomology as @xmath956 . the case that @xmath1 and @xmath614 are trivial gives the complexified multiplihedron @xmath964 constructed in section [ ziltener ] .
indeed , each irreducible component @xmath386 with finite or non - zero scaling is equipped with a isomorphism with the affine line , unique up to translation , and therefore a scaling @xmath535 . thus the underlying curve of any affine vortex is automatically a ( possibly unstable ) scaled affine curve in the sense of section [ ziltener ] .
let @xmath851 denote the moduli space of isomorphism classes of finite energy @xmath22-marked vortices on @xmath72 ( the additional marking at infinity ) with values in @xmath1 . by combining the sequential compactness theorem [ affinecompact ] with local distance functions as in ,
one has : ( properness of the moduli space of affine vortices ) suppose that @xmath1 is a hamiltonian @xmath614-manifold with proper moment map convex at infinity , equipped with an invariant compatible almost complex structure .
@xmath956 is hausdorff and the energy map @xmath965 is proper . in order to understand bubbling in the infinite area limit we briefly review the definition of a stable maps for orbifold targets , studied in chen - ruan @xcite , which appear as bubbles in the definition of certain gauged maps .
the corresponding algebraic theory by abramovich - graber - vistoli @xcite will be reviewed in ( * ? ? ?
* section 5 ) . recall that 1 . an _ orbifold structure _ on a topological space
@xmath8 is a proper tale groupoid together with a homeomorphism of the space of objects to @xmath8 ; 2 .
an _ nodal orbifold structure _ on a nodal complex curve @xmath8 is an orbifold structure on its normalization , such that the automorphism group of any nodal point is independent of the choice of irreducible component containing it .
3 . a nodal orbifold structure is a _ twisting _
if each point with non - trivial automorphism is either a node or marking , and for each node , the orbifold charts satisfy the following _ balanced condition _ :
the chart on one side of the node is of the form @xmath966 for some @xmath51 where @xmath50 acts on @xmath967 a neighborhood of @xmath23 , while the orbifold structure on the other side is is @xmath968 with the conjugate action by @xmath969 .
4 . let @xmath336 be a compact symplectic orbifold equipped with a compatible almost complex structure @xmath227 .
a _ stable map _ to @xmath336 is a complex nodal curve @xmath8 equipped with a twisting and a representable pseudoholomorphic orbifold map @xmath970 .
representability means that the map @xmath9 is smooth after passing to tale cover in @xmath336 : if @xmath971 and @xmath972 has automorphism group @xmath973 , then there exists an orbifold chart @xmath974 for @xmath336 near @xmath975 so that the groupoid fiber product @xmath976 is an orbifold chart for @xmath742 and @xmath9 has a smooth local lift @xmath977 to @xmath978 given by projection on the second factor .
in particular , @xmath979 injects into @xmath973 . the notion of gromov convergence of twisted pseudoholomorphic maps which generalizes gromov convergence of stable pseudoholomorphic maps .
let @xmath980 denote the space of isomorphism classes of connected , genus @xmath160 , @xmath22-marked stable maps to @xmath336 .
[ orbcompact ] ( properness of moduli spaces of stable maps to orbifolds ) let @xmath336 be a compact symplectic orbifold equipped with a compatible almost complex structure .
@xmath980 is hausdorff and the energy map @xmath981 is proper .
the adiabatic limit theorem [ largearea ] will be proved by studying a moduli space of curves with _ varying scaling _ which interpolates between the moduli space of vortices for a fixed area form @xmath982 and the infinite area limit .
more precisely , theorem [ largearea ] will follow from a divisor class relation in the source moduli space constructed in section [ compose ] , just as associativity of the quantum product follows from a divisor class relation in the moduli space of stable curves .
first we construct the objects that appear in the infinite area limit .
let @xmath1 be a hamiltonian @xmath614-manifold with proper moment map , so that @xmath785 is a locally free quotient and so an orbifold . 1 .
( stable map to the quotient ) a stable @xmath51-marked pseudoholomorphic map @xmath983 of class @xmath984,d_0)$ ] for some @xmath985 ; 2 .
( affine vortex bubbles ) for each marking @xmath986 a stable @xmath987-marked affine vortex @xmath988 with markings @xmath989 and orbifold structure of @xmath162 at the point at infinity @xmath990 matching the orbifold structure of @xmath284 at @xmath991 , so that the union @xmath992 is a balanced orbifold curve ; 3 . ( matching condition ) the value @xmath993 in @xmath785 , where @xmath994 is the projection on the second factor .
an _ isomorphism _ of nodal infinite area vortices is a combination of an automorphism of the underlying curves intertwining the scalings and a bundle isomorphism on the affine vortex bubbles , intertwining the markings , scalings , connections and maps .
we denote by @xmath8 the nodal curve obtained by gluing together the curves @xmath284 and @xmath386 at @xmath995 ; the matching condition means that the orbifold structures glue together to a twisting of @xmath8 .
removal of singularities for vortices in remark [ extends ] implies that orbi - bundles @xmath996 defined by the limit of the connections at infinity glue together with the bundle @xmath997 given by pull - back of @xmath810 , to give a bundle @xmath13 .
the matching condition then implies that the section @xmath998 over @xmath284 and @xmath75 over @xmath162 glue together to a section of @xmath635 over @xmath8 with fairly weak regularity properties at the nodes with infinite scaling , so that @xmath9 takes values in the zero level set @xmath802 on the subset of @xmath8 with infinite scaling .
an isomorphism of nodal symplectic vortices is then an automorphism of the domain intertwining the connections ( singular at the nodes with infinite scaling ) , maps , scalings , and markings .
let @xmath999 denote the moduli space of isomorphism classes of nodal infinite - area vortices . from the description above , @xmath999 is the union of fiber products @xmath1000 over unordered partitions @xmath1001 $ ] of @xmath576 . 1 .
( two - form corresponding to a scaling ) any scaled curve @xmath1003 in the sense of definition [ scaledcurves ] defines a volume form @xmath1004 as follows . 1 .
if @xmath39 is finite on the principal component , then @xmath1005 is @xmath1006 times the form @xmath982 on the principal component , and on the bubble component @xmath1005 vanishes .
2 . if @xmath39 is infinite on the principal component , then @xmath1005 is equal to @xmath1007 on each bubble tree .
2 . ( scaled vortices ) a _ marked scaled vortex _ with domain @xmath8 and target @xmath1 consists of a marked scaled curve @xmath1008 together with a vortex on @xmath280 corresponding to the area form @xmath1005 . that is , 1 .
if @xmath39 is finite on the principal component , then a vortex @xmath1009 on the principal component corresponding to the area form @xmath1010 and a collection of sphere bubbles @xmath1011 for each component @xmath1012 ; 2 .
if @xmath39 is infinite on the principal component , then a holomorphic map @xmath1013 on the principal component @xmath284 and a collection of nodal affine vortices on the bubble trees attached to @xmath284 ; + both should satisfy natural matching conditions at the nodes .
a marked scaled vortex is _ polystable _ if each non - principal component with finite , non - zero scaling ( resp .
zero or infinite scaling ) has at least @xmath255 ( resp .
@xmath156 ) special points .
the notion of _ isomorphism _ of marked scaled vortices , using gauge transformations and automorphisms of the domain , is left to the reader .
a marked scaled vortex is _ stable _ if it is polystable and has no infinitesimal automorphisms .
we denote by @xmath1014 the moduli space of isomorphism classes of marked scaled vortices , and by @xmath1015 the component with homology class @xmath770 .
the notion of convergence of vortices extends naturally to the case of varying scaling , and defines a topology on @xmath1014 .
[ scaledproper ] ( properness of the moduli space of scaled symplectic vortices ) let @xmath1 be a hamiltonian @xmath614-manifold with proper moment map convex at infinity equipped with a compatible almost complex structure , such that @xmath785 is a locally free quotient .
then @xmath1014 is hausdorff and the energy map @xmath1016 is proper .
[ bconnect ] for every constant @xmath1017 , there exist constants @xmath758 such that if @xmath653 is a vortex on the cylinder @xmath1018 \times s^1 $ ] with respect to an admissible area form as in with respect to some constant @xmath1019 . if @xmath1020 , then the energy density @xmath917 satisfying for @xmath1021 \times s^1 $ ] @xmath1022 we assume that @xmath1023 are as in the statement of the lemma . as in ziltener @xcite ,
the assumptions imply that for @xmath1024 sufficiently small , the invariant symplectic action @xmath1025 in @xcite is well - defined and satisfies the energy - action inequality @xmath1026 \times s^1 } ) = - { { \mathcal{a}}}((a , u)(s ' , \cdot ) ) + { { \mathcal{a}}}((a , u)(s , \cdot ) ) .\ ] ] then as in @xcite , @xmath1027 \times s^1 } ) \leq -(2\pi / a - { \epsilon } ) e(s)\ ] ] which implies the necessary exponential decay for the energy .
the exponential decay for the energy density follows from the point - wise estimate in ( * ? ? ?
* lemma 3.3 ) : if @xmath653 has sufficiently small energy on a ball of radius @xmath1028 around @xmath1029 contained in @xmath1030 \times s^1 $ ] then the energy density satisfies @xmath1031 note that the statement of lemma ( * ? ? ?
* lemma 3.3 ) is only for a fixed vortex on a half - cylinder , on a sufficiently small neighborhood of infinity , but an examination of the short proof shows that it suffices to have sufficiently small energy on the ball .
we claim that any sequence of vortices with varying scaling with bounded energy has a convergent subsequence , after passing to a limit . for sequences with scaling
bounded away from infinity , this follows from ott @xcite . for scalings approaching infinity ,
one obtains convergence in the absence of bubbling by the arguments of gaio - salamon @xcite .
they also show that bubbles in @xmath785 or @xmath1 or affine vortices satisfy energy quantization and so bubbling happens only at finitely many points .
thus one obtains , on the complement of a finite set in @xmath8 , an infinite area vortex and by proposition [ bij ] a stable pseudoholomorphic map to @xmath1032 . by removal of singularities for orbifold stable maps ,
one obtains a stable map to @xmath1033 .
the annulus lemma in parts ( c , d ) of theorem [ annlem2 ] , and the extension described in lemma [ bconnect ] , implies that bubbles in @xmath1 connect with the affine vortices . the standard soft rescaling argument
( see for statement .
let the closed sets be those for which any convergent sequence has a limit . that this defines a topology , and that this topology is hausdorff , uses the construction of local distance functions as in which will be left to the reader . to prove the adiabatic limit theorem [ largearea ]
, we will require that @xmath653 is a vortex with respect to the area form @xmath1034 for @xmath1035 .
this is because without this adjustment the moduli space @xmath1014 is almost never ( even virtually ) smooth , because of reducible pairs @xmath653 , and so can not be expected to yield invariants relating gauged gromov - witten invariants with gromov - witten invariants of the quotient . for a suitable `` master space '' relating the invariants for different @xmath122 ,
see @xcite .
e. arbarello , m. cornalba , and p. a. griffiths . , volume 268 of _ grundlehren der mathematischen wissenschaften [ fundamental principles of mathematical sciences]_. springer , heidelberg , 2011 . with a contribution by joseph daniel harris .
j. li and g. tian .
virtual moduli cycles and gromov - witten invariants of general symplectic manifolds . in _ topics in symplectic @xmath454-manifolds ( irvine ,
ca , 1996 ) _ , first int .
press lect .
, i , pages 4783 .
internat . press , cambridge ,
ma , 1998 .
m. markl .
simplex , associahedron , and cyclohedron . in _
higher homotopy structures in topology and mathematical physics ( poughkeepsie , ny , 1996 ) _ , volume 227 of _ contemp .
_ , pages 235265 .
soc . , providence , ri , 1999 .
d. mumford .
boundary points of moduli schemes . in _
lecture notes prepared in connection with the seminars held at the summer institute on algebraic geometry , whitney estate , woods hole , massachusetts , july 6 july 31 , 1964_. http://www.jmilne.org/math/documents/woodshole.pdf[woodshole.pdf ] . | this is the first in a sequence of papers in which we construct a quantum version of the kirwan map from the equivariant quantum cohomology @xmath0 of a smooth polarized complex projective variety @xmath1 with the action of a connected complex reductive group @xmath2 to the orbifold quantum cohomology @xmath3 of its geometric invariant theory quotient @xmath4 , and prove that it intertwines the genus zero gauged gromov - witten potential of @xmath1 with the genus zero gromov - witten graph potential of @xmath4 . in this part
we introduce the moduli spaces used in the construction of the quantum kirwan morphism . |
knowing the redshifts of extragalactic objects is vital for understanding their physical properties as well as for statistical cosmology measurements .
the primary method employed by researchers to measure galaxy redshifts is spectroscopy ; however , spectroscopy is available for only a small fraction of all imaged galaxies , on the order of 1% @xcite .
this low percentage has necessitated a search for alternative methods of redshift estimation .
much progress has been made with photometric redshift estimation methods , which primarily rely on the colors of galaxies to estimate redshifts @xcite . while such methods are used frequently , they face inherent limitations , resulting in redshift estimates with large associated error values .
consequently , methods of redshift estimation using the angular clustering of galaxies have been proposed and explored @xcite .
because these methods require only the angular positions of galaxies , they are considered distinct from photometric redshift methods per the working definition argued by @xcite .
the `` clustering - based '' redshift estimation methods rely on the spatial correlations and directly use the angular two - point correlation function @xmath0 , which expresses the excess probability of finding a galaxy at an angular separation @xmath1 from another galaxy .
while these methods are promising , they propose only redshift distributions and thus are agnostic to the redshifts of specific galaxies .
therefore , a natural question arising from clustering - based redshift estimation is whether galaxy clustering can be used not only to produce redshift distributions of galaxy samples but also to determine the redshifts of individual galaxies . in this paper
, we propose a new method that uses combinatorial optimization techniques applied to the angular two - point correlation function to divide galaxy samples into separate redshift bins , thereby constraining the redshifts of individual galaxies .
creating thin slices not only is possible but potentially makes the optimization faster .
ultimately , the final method is expected to balance the statistical noise and the optimization cost . here
we explore the simplest case .
we aim to partition a set of ( simulated ) galaxies into two subsamples such that their correlation functions match the observations
. this case would be particularly useful in creating a hard boundary between overlapping photo - z redshift bins .
if it works , the optimization can be repeated iteratively in order to separate a galaxy sample into narrow redshift slices .
of course , even this simplest case presents computational challenges .
[ line - through]*naively , for a sample of size @xmath2 , @xmath3 possible partitions would have to be considered in order to find the optimal partition , which is prohibitively expensive even for dozens of objects .
* we present a formalism using integer linear programming that , when implemented , makes this optimization tractable for statistically relevant sample sizes , . in section [ sec : formalism ]
, we provide a detailed description of our method , including a derivation of the relevant formalism . in section [ sec : results ] , we present results from in section [ sec : applicability ] , we discuss the applicability of this method .
before introducing our formalism , let us introduce the notation that will be used throughout the rest of this paper .
let @xmath4 and @xmath5 be two datasets with the same sky coverage , and let @xmath6 be a random dataset with the same sky coverage as @xmath4 and @xmath5 . in accordance with @xcite we define @xmath7 as the set of all unordered pairs of galaxies in @xmath4 , and we define @xmath8 as the set of all unordered pairs of galaxies such that one galaxy is from @xmath4 and the other is from @xmath5 .
we also define @xmath9 and @xmath10 analogously .
because the correlation function is estimated over a set of @xmath2 bins ( i.e. , intervals ) of angular separation , we must introduce notation related to pair counts within these bins .
we define @xmath11 as the set of unordered data - data pairs such that the angular separation between the members of the pair is in bin @xmath12 , @xmath13 .
the corresponding terms for @xmath14 , @xmath10 , etc . , are defined analogously .
lastly , we define @xmath15 as the size of @xmath4 , @xmath16 as the size of @xmath11 , etc . in this paper , we develop our formalism using two estimators of correlation functions given two samples with the same sky coverage : the natural estimator and the @xcite estimator .
the natural estimator is the least expensive of the correlation function estimators to compute and hence expected to yield the simplest optimization formulas . on the other hand ,
the landy - szalay estimator has been shown to be the most accurate estimator by @xcite . using the natural estimator , the cross - correlation and autocorrelation function estimates of the two samples
@xmath4 and @xmath5 in bin @xmath12 are given by : @xmath17 @xmath18 @xmath19 in the main body of the paper , we focus on these equations but provide the alternative method for the landy - szalay estimator in appendix [ sec : ls ] .
one of the most powerful tools in discrete optimization is _ integer linear programming ( ilp)_. _ linear programming _ is an optimization model in which one seeks to optimize a linear function of finitely many variables , subject to linear inequality constraints on the variables .
integer linear programming generalizes this by only admitting solutions where some ( or all ) of the variables are constrained to take only integer values .
such models are an indispensable tool for optimizing over discrete solution spaces , and state - of - the - art software codes have been developed to handle numerous large scale problems arising in technological , scientific , and economic applications . in order to build an integer linear programming model ,
we must build a model consisting of a cost function and constraints , where our cost function and constraints are constructed such that the minimization of our cost function yields the optimal solution to our problem .
we define our optimal solution as a partition of our sample @xmath20 into two subsamples @xmath21 and its complement @xmath22 such that : @xmath23\ ] ] is minimized , where @xmath24 , @xmath25 , and @xmath26 are our correlation function estimates in bin @xmath12 , as calculated using our choice of estimator , and @xmath27 and @xmath28 are target values for @xmath24 , @xmath25 , and @xmath26 , respectively .
informally , this function is minimized when @xmath29 , @xmath30 , and @xmath31 are pulled as close to our target values @xmath27 and @xmath28 as possible , across all bins @xmath12 . the purpose of the following formalism is to translate equation [ eq : tot ] into an integer linear programming model .
the construction of such a model requires translating all unknowns into variables , creating a cost function using a linear combination of these variables , and adding constraints to the model to enforce certain relationships between the variables .
we must also specify which variables are integer or continuous . in current software
, one can even specify if an integer variable is binary , i.e. , takes only values @xmath32 or @xmath33 .
our integer variables will , in fact , be binary variables to accommodate the problem of classifying objects into two redshift slices .
to simplify the optimization further , we fix the size of @xmath21 and its complement , @xmath34 and @xmath35 , to prevent runaway solutions in which one subsample contains a large majority of the galaxies in @xmath20 .
moreover , this enables us to keep the cost function and constraints linear .
consequently , we treat @xmath34 and @xmath35 as constants throughout .
this , however , poses no real threat to generality because often good initial estimates are available , and further optimization can be performed along the size dimension if needed .
we begin by defining variables for each of our galaxies .
we introduce the binary variable @xmath36 that encodes whether a galaxy @xmath37 is a member of @xmath21 or @xmath22 : @xmath38 these variables serve as the bridge between the cost function and the partitioning of @xmath20 : we will construct the cost function in such a way that minimizing it sets each @xmath36 to either 0 or 1 and thus assigns each galaxy to @xmath21 or @xmath22 according to the optimal partition .
we now add a constraint to our model in order to enforce that @xmath34 must be fixed to a pre - determined positive integer by using the fact that @xmath34 is precisely the sum of @xmath39 that evaluate to @xmath33 : @xmath40 thus , we have fixed @xmath34 , and because , where @xmath41 is fixed , we have also fixed @xmath35 .
next , we introduce variables for unordered pairs of galaxies in @xmath20 that encode whether the galaxies in each pair are from the different subsamples . for each unordered pair of galaxies
, we define the binary variable @xmath42 as follows : @xmath43 where @xmath42 is symmetric in @xmath37 and @xmath44 .
significantly , @xmath42 can be expressed in terms of the boolean ( xor hereafter ) of the @xmath36 and @xmath45 variables , @xmath46 where each xor is encoded through four linear constraints of @xmath47 and @xmath42 that we add to our model as @xmath48 we reiterate that these constraints establish the relationship between @xmath49 s and @xmath50 s , as given in their definitions in and . before proceeding , we introduce more notation - related summations over pair counts .
we use the summation notation @xmath51 to denote summing over all unordered pairs of galaxies in @xmath20 such that the angular separation between @xmath37 and @xmath44 falls into bin @xmath12 .
we begin by translating the natural estimator @xmath52 to its linear programming equivalent in this section and generalize to the landy - szalay estimator @xmath53 in the appendix . using the natural estimator
, we first translate @xmath31 into its cost function equivalent , @xmath54 . for cross - correlation
, we seek to minimize : @xmath55 where @xmath26 is an estimate of cross - correlation in bin @xmath12 .
we can express @xmath26 in terms of previously - defined quantities : @xmath56 because @xmath6 is fixed , @xmath57 and @xmath58 are constants .
furthermore , @xmath34 and @xmath35 are fixed .
thus , we can combine these constants into a single weight for each bin : @xmath59 we therefore seek to minimize @xmath60 where @xmath61 is the only non - constant term within the minimization for each bin @xmath12 .
we can now reformulate this expression using our previously - defined binary variables @xmath42 .
@xmath62 is precisely equal to the number of unordered pairs @xmath63 of galaxies in @xmath20 such that @xmath37 and @xmath44 are in different subsamples and the angular separation between @xmath37 and @xmath44 falls into bin @xmath12 ; thus , @xmath64 because we are summing over all unordered pairs in @xmath65 , as opposed to just unordered pairs in @xmath66 , we have eliminated any dependence on the partitioning of @xmath20 except in the variables themselves .
the expression that we seek to minimize for cross - correlation optimization now simplifies to : @xmath67 because each @xmath42 is the xor of @xmath36 and @xmath45 , we have expressed cross - correlation optimization entirely in terms of binary variables for each galaxy , and minimization of the above expression will assign each galaxy to either @xmath21 or @xmath22 according to the optimal partition of @xmath20 .
the expression in is not a linear function of the @xmath42 variables due to the absolute value function .
however , this expression can be modeled by a linear function by introducing auxiliary continuous variables @xmath68 for each bin @xmath12 and relating them to the @xmath42 variables as follows .
@xmath69 where for each @xmath68 , we add the following two constraints : @xmath70 we have now incorporated cross - correlation optimization into our linear programming model .
the key insight here is that any solution that minimizes the expression in must satisfy one of the inequalities in at equality depending on which left hand side is larger , and thus @xmath71 in any optimum solution minimizing .
we mention here the conscious choice of using the @xmath72 norm in , as opposed to the @xmath73 norm : the @xmath72 norm leads to a formulation with a linear objective like and linear constraints like , as opposed to the @xmath73 norm which would give a convex quadratic objective . as per folk wisdom in integer programming ,
we prefer the linear formulation , and hence we go with the @xmath72 norm .
next , we formalize autocorrelation optimization using the natural estimator . in this paper , we formalize two different approaches : combining the autocorrelations of @xmath21 and @xmath22 in such a way that the autocorrelation target values @xmath74 and @xmath75 of @xmath21 and @xmath22 , respectively , are set to be equal , and implementing separate target values for the autocorrelation of @xmath21 and @xmath22 , thereby allowing @xmath74 and @xmath75 to be potentially distinct for any bin @xmath12 .
although the latter method is advantageous because it allows for independent target values , the downside is that one needs more variables in the integer linear program to model this , compared to the former method in which no new variables need to be introduced into our model .
we introduce the former method next and introduce the latter method in the appendix in [ sec : independent ] . to derive a parameterized autocorrelation for the two samples , let us consider the variable @xmath76 defined as @xmath77 from the definition we can see that . although @xmath76 is agnostic as to which sample @xmath37 and @xmath44 belong
, it does encode whether the pair contributes to an autocorrelation calculation .
we can naturally extend the notion of autocorrelation for @xmath21 and @xmath22 into a combined autocorrelation given by : @xmath78 where @xmath79 is the weighted average of @xmath24 and @xmath25 : @xmath80 with @xmath81 in this combined autocorrelation model , we seek to minimize @xmath82 where @xmath74 is the target value for the combined autocorrelation in bin @xmath12 .
we introduce the weight @xmath83 to replace constants in @xmath84 : @xmath85 furthermore , we can express entirely in terms of variables that have already been introduced because this sum is precisely equal to the sum of unordered pairs @xmath86 of galaxies in @xmath20 such that @xmath37 and @xmath44 are in the same subsample and the angular separation between @xmath37 and @xmath44 falls into bin @xmath12 : @xmath87 our expression takes the form : @xmath88 - ( 1 + \alpha_i ) \bigg|\ ] ] as with cross - correlation optimization , the final step is to eliminate absolute values from the cost function . for each bin @xmath12 , we add a continuous variable @xmath89 .
the portion of the cost function @xmath90 corresponding to combined autocorrelation optimization takes its final form : @xmath91 where for each @xmath89 , we add the following two constraints : @xmath92 - ( 1 + \alpha_i ) & \le \psi_i\\ \bigg [ b_i \sum_{(u ,
v ) \in vv_i } ( 1 - y_{uv } ) \bigg ] - ( 1 + \alpha_i ) & \le - \psi_i \end{split}\ ] ] we can now express our entire model using combined autocorrelation and the natural estimator .
the model consists of the cost function : @xmath93 and all associated constraints .
this model can be generalized to the landy - szalay estimator by modifying the cost function slightly ; we provide the details in the appendix in [ sec : ls ] .
furthermore , in all of the above formalism , we have used a uniform weighting across all bins for cross - correlation and autocorrelation in the cost function for notational simplicity .
we note that one can easily weight each bin individually .
to test the effectiveness of the integer linear programming method described in section [ sec : formalism ] , we used the python interface for gurobi , a solver available for both academic and commercial use .
we ran all tests on a dell poweredge r815 machine with 512 gb of ram and 4 amd opteron 6272 processors .
the machine ran scientific linux 7 , python 2.7.5 , and gurobi 6.0.4 .
it is worth noting that none of the following tests required gurobi to use more than 1% of ram at any given time ; therefore , our machine s specifications far exceed the specifications necessary to reproduce the following results .
for the two datasets @xmath21 and @xmath22 , we used two mock catalogs , each 1 square degree in size and each consisting of 10,000 galaxies , generated [ line - through]*by sbastien heinis * using a cox point process , as described in heinis et al .
the samples were produced using different random seeds , and consequently , they behaved as uncorrelated samples , meaning that the theoretical cross - correlation was 0 . in both samples , we selected the thin redshift cut of , the densest redshift slice in both samples , in order to select samples with strong angular autocorrelation signals .
these cuts left slightly over 1000 galaxies per catalog .
we then randomly selected 2000 galaxies in total from these two samples , yielding @xmath21 and @xmath22 such that , for our random catalog , we generated a 1 square degree poisson sample consisting of 200,000 galaxies .
all bins for cross - correlation and autocorrelation were given equal weights in the cost function according to the formalism derived in section [ sec : formalism ] .
furthermore , we chose to use binning schemes with an equal number of galaxy pairs per bin across all bins . with this choice ,
the poisson noise was equal across all bins , unlike binning schemes set by uniform increments in angular separation .
test the effectiveness of the optimization itself , we fed gurobi ground - truth target values by pre - computing all @xmath74 s and @xmath28 s using the real partition of @xmath20 into @xmath21 and @xmath22 ; we also fixed @xmath34 and @xmath35 according to equation [ eq : ssize ] using the real partition of v. given the ground - truth as target values , we then tested 1 . the time required for gurobi to complete the optimization 2 .
the fraction of galaxies assigned correctly accordingly , these are the main metrics by which we compare different optimization runs in this section . in the limit of sufficiently few pairs per bin , the optimization is computationally feasible and recovers the ground - truth solution exactly .
for instance , with a binning scheme of @xmath94 pairs per bin and @xmath95 bins , the optimization recovers the ground - truth partition in @xmath96 seconds .
however , in this case of only 5 pairs per bin , the optimization almost certainly couples to the noise of the ground - truth correlation function target values .
this motivated an analysis of a binning scheme of 100 pairs per bin , which improves upon the poisson noise per bin while still retaining resolution in the autocorrelation signal at short scales by having sufficiently narrow bins . in order to reduce computational load and make the optimization feasible , we fixed a fraction of the total galaxy population before beginning the optimization by adding extra constraints to our model that if and if for a fraction of galaxies ; including a fixed sample is reasonable because a spectroscopic sample could be used as a fixed population in an application of this method to real data . in order to explore the optimization with these settings , we ran a series of tests using gurobi with @xmath97 of galaxies fixed and a binning scheme of 100 pairs per bin . . three distinct regimes
the first regime occurs in the limit of few bins , or short maximum angular scales , in which there are many solutions with 0 cost .
because there are so many optimal solutions , gurobi is capable of finding one such optimal solution in a short amount of time ; however , for this same reason , this solution behaves only slightly better than random in assigning the unfixed galaxies correctly .
it is worth noting that even in this regime , the amount of time required to complete optimization increases exponentially as a function of number of bins .
this increase in runtime is expected because increasing the number of bins increases the number of target values and thus decreases the number of solutions with 0 cost .
the second regime occurs in the limit of many bins , in which the optimization is communicated a sufficient amount of information about the ground - truth solution through the @xmath74 s and @xmath28 s that it is able to find the ground - truth solution in of order a minute .
consequently , gurobi assigns 100% of the unfixed galaxies correctly .
the third regime is the peak in the amount of time required for the optimization to complete , located between the other two regimes .
as seen in , the full features of the peak can not be determined due to the exponential growth in runtime .
the points correspond to optimization runs that were terminated before completion after @xmath98 seconds , the maximum allowed time in our study ; the fraction of galaxies correctly assigned corresponds to that of the optimal solution found before termination .
this peak is the result of a trade - off between the other two regimes : there are relatively few solutions with 0 cost , and they can not be found quickly by gurobi ; however , there are not enough bin target values to constrain the ground - truth solution immediately . in figure
[ fig:100leftright ] , for a binning scheme with 200 pairs per bin .
the same three characteristic regimes are present ; however , the second and third regimes begin at shorter maximum angular binning scales for the binning scheme of 100 pairs per bin than for the binning scheme of 200 pairs per bin .
this is a consequence of the fact that for a given maximum angular binning scale , the binning scheme of 100 pairs per bin is fed twice the number of @xmath74 s and @xmath28 s than the binning scheme of 200 pairs per bin and thus has twice the information .
the results presented in figure [ fig:100leftright ] are highly sensitive to the fraction of galaxies fixed before optimization .
for example , in the regime of many bins , the optimization requires an order of magnitude more time to complete the optimization for 75% of galaxies fixed than for 80% of galaxies fixed ; this optimization does not complete within @xmath98 seconds for 70% of galaxies fixed .
reducing the total number of galaxies to @xmath99 , the optimization requires a lower percentage of galaxies to be fixed in order to recover the ground truth solution in the limit of many bins .
fixing 65% of galaxies and using 100 edges per bin and 600 bins , the optimization completes in 195 seconds . lowering the percentage fixed to 60%
, the optimization completes in 695 seconds .
thus , reducing the total number of galaxies to @xmath99 still allows for @xmath100 unfixed galaxies to be assigned , confirming that the percentage of fixed galaxies necessary for the optimization to complete within a given time frame is largely dependent on the total number of galaxies in @xmath20 .
the results presented in section [ sec : results ] reveal that in the appropriate regimes , the optimization is computationally feasible when ground - truth values are fed in as the target values . in any real application of this method , the ground - truth values of the @xmath74 s and @xmath28 s would only be known approximately , and fiducial values are good approximations in the case of many pairs in the bins .
the determination of the appropriate fiducial values is a question of physics rather than of linear programming , and because we have chosen to test only the optimization itself in this paper , so far we have omitted the exploration of the effects of inputting fiducial values on the optimization .
we test the optimization s response to inexact target values by perturbing the @xmath74 s toward values taken from a power law fit of the combined autocorrelation function and perturbing the @xmath28 s toward @xmath32 , the expected cross - correlation .
we accomplish this by using interpolation and setting the target values by varying the interpolation parameter @xmath101 $ ] according to the following equations : @xmath102\ ] ] and @xmath103 where @xmath104 is the ground - truth combined autocorrelation value in bin @xmath12 , @xmath105 is the ground - truth cross - correlation value in bin @xmath12 , and @xmath106 is the value of the power - law fit in bin @xmath12 .
as seen in figure [ fig : corrfn ] .
this power law fit was used only for the purposes of perturbing the ground - truth combined autocorrelation function by small amounts , as seen in figure [ fig : interpolation ] . in figure
[ fig : interpolation ] , we present the time required for the optimization to complete when varying @xmath107 up to 0.04 for 80% of galaxies fixed , 200 pairs per bin , and 300 bins . by varying @xmath107 only by small amounts ,
we lessen the impact of an incorrect power law fit .
furthermore , these binning settings were chosen because the optimization recovered the exact ground - truth partition in 44 seconds given these settings and ground - truth target values .
for all runs in this figure , the optimization exactly recovered the ground - truth solution .
the fact that the optimization still completes and recovers the real partition when given inexact target values for for , the optimization does not complete within @xmath98 seconds , the observed phenomenon of an increase in runtime with @xmath107 is possibly due to the fact that the model might not have a perfect solution and the cost function could become shallow , which leads to a large number of nearly optimal " solutions .
these solutions must be pruned by the solver to find the global minimum conclusively .
however , the pruning methodologies of ip solvers often need to spend a significant amount of time to discard the
nearly optimal " solutions , even though they have arrived at a stage of the optimization where all the solutions that are being considered have values that are very close to the true optimal value . to finally discover the true optimal value by weeding through the large number of `` nearly optimal '' ones occupies the bulk of the time for
the solver in such situations ( see chapter 2 of @xcite for a discussion of these pruning strategies for integer linear programming ) . by interpolating a power law fit
, we have in effect pulled the @xmath74 target values to a slightly less - noisy correlation function . in order to resolve fully the question of whether the optimization is finding the ground - truth solution by coupling to noise
, @xmath107 would have to be driven closer to 1 ; of course , in doing so , the target values would become more dependent on the fiducial values , and physically - correct fiducial values would have to be chosen , .
however , a visual representation of the noise in each bin of the ground - truth combined autocorrelation function is nonetheless instructive and can be seen in figure [ fig : corrfn ] .
our ultimate goal is to use this procedure in conjunction with photometric redshift results to sort galaxies into thin redshift slices using only photometric data .
the true redshift distribution of photo - z bins will have overlapping tails due to the uncertainties in the estimation . starting with designations from the photo - z catalog
, one can apply our procedure to create a sharper boundary between the bins .
repeating this procedure will yield much improved redshifts . .
we have presented a novel integer linear programming method that enables a galaxy sample to be partitioned into two subsamples such that the angular two - point cross- and autocorrelations of the subsamples are optimized to pre - determined target values .
our approach is the first application of integer linear programming to , which is expected to find other applications as we explore large statistical samples .
we tested this optimization method using mock catalogs and gurobi , an optimization solver , and verified that this optimization technique is not only feasible in certain regimes but also provides good solutions , .
this is due to the formulation of the problem using only linear equations .
we explored the applicability of this method and have described how it could be used in the future to estimate the redshifts of individual galaxies using only their celestial coordinates .
evidently , much more remains to be explored with the applicability of this method , a significant portion of which relates to fiducial values .
for instance , the question of the extent to which the optimization is coupling to noise in the correlation function can only be resolved by analyzing runs with fiducial target values .
furthermore , much would be learned from re - generating figure [ fig:100leftright ] using fiducial values as target values once the best method for selecting fiducial values is identified .
other variables such as the optimal binning scheme and maximum angular binning separation for the autocorrelation and cross - correlation when using fiducial values would also have to be explored . ideally ,
for real application , the number of galaxies per sample could be pushed higher , and the fraction of galaxies fixed could be pushed lower ; given the rapid improvement in linear programming algorithms and the increase in computing power , these limitations may very well resolve themselves with time .
additions to the outlined method could be explored in order to improve runtimes and accuracy .
some of these are mathematical and others physical issues .
\(1 ) in a real application of this method , the optimization could be terminated when the relative gap between the upper and lower bounds of the optimization is below a threshold value , instead of forcing this gap to reach zero in order for the optimization to complete and find the true _ mathematically _ optimal solution . in this case , the solver may report a `` nearly optimal '' solution of the optimization model , as opposed to the true _ mathematical _ optimum of the model .
nevertheless , we feel that from a physical standpoint , the `` nearly optimal '' solutions may be more meaningful than the mathematical optimal which could be influenced by noise .
such a strategy would also make sense when we use fiducial values instead of values from a simulation and would help to deal with the explosion in running time as discussed in section [ sec : toward_fiducial ] .
\(2 ) other than through the @xmath74 s , the power law nature of the autocorrelation function is never explicitly leveraged ; it could potentially be exploited in a greedy algorithm , for example , by fixing pairs of shortest angular separations to the same subsample before beginning optimization .
the autocorrelation signal could also potentially be leveraged to a greater extent by using a binning scheme set by uniform increments in angular binning separation , which could provide more bins at the shortest angular scales .
\(3 ) it is possible that subsampling the pairs in each bin could decrease runtime without sacrificing accuracy .
considering that the true underlying variables are the classes to which the objects belong ( @xmath36 ) , a subset of the pairs ( @xmath42 ) could provide enough constraints at a reduced computational cost .
the sampling , however , will probably have to be carefully constructed to optimize performance .
the authors would like to thank sbastien heinis for providing mock catalogs and brice mnard for helpful discussions .
lance joseph helped with computational resources .
b.l . was supported by the 2015 herchel smith - harvard undergraduate science research fellowship .
a.b . gratefully acknowledges partial support from nsf grant cmmi1452820 .
we formalize the independent autocorrelation method for the natural estimator and provide the formalism for the landy - szalay estimator .
the mathematical notation for these are somewhat more complicated but the complexity of the algorithm does not increase .
here , we introduce the formalism for autocorrelation optimization with the natural estimator in which target values @xmath74 and @xmath75 can be set independently , as opposed to the combined autocorrelation method described in [ sec : combined ] .
the independence of @xmath74 and @xmath75 comes at the expense of model complexity : we must introduce the variables @xmath108 and @xmath109 and associated constraints for each unordered pair , where neither @xmath108 nor @xmath109 can be expressed in terms of @xmath42 .
therefore , this method effectively triples the number of variables in our model in comparison to the combined autocorrelation model .
we designate the cost function equivalent of the autocorrelation of @xmath21 and @xmath22 using the natural estimator as @xmath110 and @xmath111 , respectively . in order to formalize the autocorrelation of @xmath21
, we introduce the variable @xmath108 . for each unordered pair , we define @xmath108 as follows : @xmath112 in addition , to formalize the autocorrelation of @xmath22 , we introduce the analogous variable @xmath109 . for each unordered pair , we define @xmath109 as follows : @xmath113 just as with @xmath50 s , we must add constraints to relate @xmath49 s to @xmath114 s and @xmath115 s : @xmath116 @xmath117 for the autocorrelation of @xmath21 , we minimize : @xmath118 and for the autocorrelation of @xmath22 , we minimize : @xmath119 thus , @xmath120 takes the final form of : @xmath121 where for each @xmath89 , we add the following two constraints : @xmath122 and @xmath123 takes the final form of : @xmath124 where for each @xmath89 , we add the following two constraints : @xmath125 in order to convert @xmath110 and @xmath126 into their final forms , we must eliminate the absolute values using the method described in sections [ sec : cross ] and [ sec : combined ] . in this formalization using independent autocorrelations for @xmath21 and @xmath22 and using the natural estimator , the full model consists of the cost function : @xmath127 and all of the associated constraints .
here , we introduce the formalism for optimization with the landy - szalay estimator . using the landy - szalay estimator , the autocorrelation and cross - correlation function estimates of two samples @xmath4 and @xmath5 in bin @xmath12 are given by : @xmath128 @xmath129 @xmath130 in building the ilp model for the natural estimator , we have already modeled the first terms in all three expressions , , and .
thus , we only need to translate the second terms in these expressions .
fortunately , they can be expressed entirely in terms of constants and @xmath36 s .
we begin with the @xmath131 term . in and
, this term in the @xmath12th bin is defined as : @xmath132 furthermore , we know that @xmath133 is defined as the number of unordered pairs such that the angular separation between @xmath37 and @xmath44 lies in bin @xmath12 . defining @xmath134 for a given as the number of ordered pairs such that and the angular separation between @xmath37 and @xmath114 lies in bin @xmath12 , we can re - express @xmath135 : @xmath136 we can in turn express this in terms of our variables @xmath36 : @xmath137 by generalizing to summing over all galaxies in @xmath20 as opposed to just galaxies in @xmath21 , we have eliminated any dependence on the partitioning of @xmath20 except in the variables themselves .
furthermore , for a given galaxy , @xmath134 is a constant that can be pre - computed before optimization .
we now define the weight @xmath138 to absorb all of these constants : @xmath139 thus , in the @xmath12th bin , the term involving @xmath131 becomes : @xmath140 we can define the term involving @xmath141 in the @xmath12th bin analogously : @xmath142 where : @xmath143 thus , referring to , , , we can now express @xmath31 , @xmath29 , and @xmath30 , respectively , in terms of our binary variables : @xmath144 where @xmath145 has been defined in equation [ eq : a ] , and @xmath146 and @xmath147 have been defined in equations [ eq : as ] and [ eq : asbar ] , respectively .
the cost function equivalents of @xmath31 , @xmath29 , and @xmath30 , given by @xmath54 , @xmath110 , and @xmath111 , respectively , can now be converted to their final forms by eliminating the absolute values using the method described in sections [ sec : cross ] and [ sec : combined ] .
thus , our full model consists of the cost function : @xmath148 and all associated constraints .
+ bentez , n. 2000 , apj , 536 , 571 benjamin , j. , van waerbeke , l. , mnard , b. , & kilbinger , m. 2010 , mnras , 408 , 1168 brammer , g. , van dokkum , p. , & coppi , p. 2008 , apj , 686 , 1503 budavri , t. , szalay , a. , connolly , a. , csabai , i. , & dickinson , m. 2000 , aj , 120 , 1588 budavri , t. , csabai , i. , szalay , a. , connolly , a. , & szokoly , g. 2001 , aj , 122 , 1163 budavri , t. 2008 , apj , 695 , 747 conforti , m. , cornujols , g. , zambelli , g. , integer programming , _ graduate texts in mathematics , springer - verlag _ , 2015 .
connolly , a. , csabai , i. , szalay , a. , koo , d. , kron , r. , & munn , j. 1995 , aj , 110 , 2655 feldmann , r. , et al .
2006 , mnras , 372 , 565 gurobi optimization , inc . , 2015 ,
gurobi optimizer reference manual , http://www.gurobi.com heinis , s. , budavri , t. , & szalay , a. 2009 , apj , 705 , 739 kerscher , m. , szapudi , i. , & szalay , a. 2000 , apj , 535 , l13 koo , d. 1985 , aj , 90 , 418 koo , d. 1999 , in asp conf .
191 , photometric redshifts and high redshift galaxies , ed .
weymann , r. , storrie - lombardi , l. , sawicki , m. , & brunner , r. ( san francisco , ca : asp ) , 3 landy , s. d. , & szalay , a. s. 1993 , apj , 412 , 64 mnard , b. , scranton , r. , schmidt , s. , morrison , c. , jeong , d. , budavri , t. , & rahman , m. 2013 , arxiv e - prints , arxiv:1303.4722 newman , j. 2008 , apj , 684 , 88 rahman , m. , mnard , b. , & scranton , r. 2015a , arxiv e - prints , arxiv:1508.03046 rahman , m. , mnard , b. , scranton , r. , schmidt , s. , & morrison , c. 2015b , mnras , 447 , 3500 rahman , m. , mendez , a. j. , mnard , b. , et al .
2016 , , 460 , 163 schmidt , s. j. , mnard , b. , scranton , r. , morrison , c. , & mcbride , c. k. 2013 , mnras , 431 , 3307 schmidt , s. j. , mnard , b. , scranton , r. , et al .
2015 , mnras , 446 , 2696 | we propose a new method of the redshifts of individual extragalactic sources based on .
techniques from integer linear programming are utilized to optimize simultaneously for the angular two - point cross- and autocorrelation functions .
our novel formalism introduced here not only transforms the otherwise hopelessly expensive , brute - force combinatorial search into a linear system with integer constraints but also is readily implementable in off - the - shelf solvers .
we adopt and use python to build the cost function dynamically .
the preliminary results on simulated data show for future applications to sky surveys by complementing and enhancing photometric redshift estimators .
our approach is the first |
starburst clusters with ages of a few million years represent unique astrophysical laboratories , as stars across the entire stellar mass range from the upper mass cut - off in the mass function down to the hydrogen burning limit ( and possibly beyond ) , and with the same metallicity and age are present in a rather homogeneous environment . as such , starburst clusters are the ideal places to study star formation and to test theories on stellar and cluster formation and evolution . unlike interacting galaxies like the antennae galaxies , where 100s of starburst clusters have been identified ( whitmore & schweizer 1995 ) , the milky way houses only a handful of starburst clusters .
starburst clusters in the antennae , however , are barely resolved , restricting us to study the integrated properties of 100,000s of stars . in the milky way , on the other hand
, starburst clusters can be resolved into 1,000s to 10,000s of stars , and the properties of each star can be derived individually . what exactly is a `` starburst '' and what defines a `` starburst cluster '' ? in general
, a starburst is a special place in space and time , where a spike in the star formation rate significantly above the `` average rate '' is observed .
the orion nebula cluster is often quoted as an example for a nearby starburst . in the following
, we are concentrating on considerably more extreme star formation events . in our ( admittedly somewhat loose ) definition , a starburst cluster is one of the most extreme star formation environments found in present - day milky way with the following properties : + a starburst cluster * contains at least several 10,000 stars * contains at least 10,000 m@xmath0 in stellar mass * houses massive stars with initial masses of @xmath1120m@xmath0 ( corresponding to an mk - type of o2v ) , width=377 ] milky way starburst clusters can be found in two different environments .
the @xmath12myr old arches cluster and the 36myr old quintuplet cluster are located near the center of the milky way with projected separations of less than 100pc from the galactic center .
the 1myr old ngc 3603 young cluster ( yc ) , and the 35myr old clusters westerlund 1 ( brandner et al .
2008 ) and 2 ( ascenso et al . 2007 ) are located in spiral arms at distances of 5 to 8kpc from the galactic center .
because of strong extinction and high stellar density in the galactic plane , our census of galactic starburst clusters is most likely incomplete , as is also high - lighted by the recent discovery of the two embedded red supergiant clusters rsgc1 and rsgc2 located in the scutum - crux spiral arm ( davies et al .
2007 ) . 5 to 10myr
ago , rsgc1 and rsgc2 would most likely have qualified as starburst clusters according to our definition .
, width=302 ] there are several advantages in studying local , and hence spatially resolved starburst clusters .
the large number of stars is crucial for a statistically sound determination of the mass function and dynamical properties of the clusters .
compared to less extreme star formation environments , starburst clusters initially house the most massive and luminous o - type main sequence stars .
uv photons from these massive stars lead to rapid photo - evaporation of any remnant circumstellar material around the low - mass members of the cluster .
this in turn brings two advantages .
first , there is very little differential extinction and ir excess .
while this is bad news for anyone looking for circumstellar disks and planets , it results in a well constrained colour - magnitude sequence for the cluster members .
secondly , the absence of circumstellar material means that non - accreting pre - main sequence tracks can be used to compare theory with observations . as ,
e.g. , shown by siess et al .
( 1997 ) or zinnecker & yorke ( 2007 ) , , the presence of on - going accretion alters pre - main sequence evolutionary tracks quite drastically ( figure [ wb_fig2 ] ) .
the following analysis is based on near - infrared observations of westerlund 1 .
seeing limited wide - field data obtained with the eso ntt and sofi cover an area of @xmath2 centered on westerlund 1 ( brandner et al .
this is supplemented by adaptive optics high - resolution imaging of the cluster center obtained with naco at the eso vlt .
, width=491 ] figure [ wb_fig3 ] shows a ks vs. h ks naco colour magnitude diagramme for a @xmath3 central field located just to the east of the center of westerlund 1 .
unrelated foreground and field stars has been statistically subtracted based on the observations of the comparison field .
the well defined cluster main - sequence , transition region and pre - main sequence are indicated .
theoretical evolutionary tracks and the derived isochrones differ in particular in their prediction of the transition region . in the present example ,
tracks by palla & stahler ( 1999 ) give a better fit to the observations than tracks by siess et al .
still , a finer mass - sampling of the ps99 tracks would be required for a more detailed calibration against the pms / ms transition region .
given a total stellar mass of at least several 10,000 solar masses , starburst clusters must have formed out of giant molecular clouds .
once the most massive stars appear on the main sequence , they rapidly ionize and disperse the remaining gas .
simulations by a variety of groups ( e.g. hills 1980 , lada et al .
1984 , geyer & burkert 2001 ) indicate that in general a star formation efficiency ( sfe ) of at least 30% is required for a stellar cluster to remain bound , though under special circumstances a sfe as low as 10% might suffice for clusters to survive for 100myr ( adams 2000 , baumgardt & kroupa 2007 ) . to answer the question if any of the local starburst clusters constitutes a proto - globular cluster , observation of the cluster kinematics are required .
thus far 1d velocity dispersions derived from radial velocity measurements of a handful of the brightest cluster members in arches ( figer et al .
2002 ) and westerlund 1 ( mengel & tacconi - garman 2007 ) have been obtained , and - assuming virial equilibrium - employed to estimate an upper limit of the total mass in each of the two clusters . 10myr . the apparent increase in cluster size as measured by the half - mass radius with increasing age is suggestive of a rapid dynamical evolution ( and dissolution in the general galactic field ) of the starburst clusters.[wb_fig4],width=491 ] figure [ wb_fig4 ] shows a picture gallery of near - infrared observations of galactic starburst clusters all plotted to the _ same _ physical scale , and ordered according to age from left to right .
only ngc 3603 yc and arches , the two youngest clusters in the sample , exhibit compact cores with half - mass radii of less than 0.5pc , whereas the already slightly more evolved westerlund 1 and quintuplet clusters have half - mass radii of 1pc .
the two recently discovered red super giant clusters ( figer et al . 2006 ; davies et al .
2007 ) with ages of around 10myr have still larger half - mass radii . for the spiral arm clusters , which are experiencing only weak tidal fields
, this could be evidence that dynamical evolution is accelerated by the gas expulsion ( see baumgardt & kroupa 2007 for recent simulations ) .
for the starburst clusters in the galactic center region , strong tidal shear could result in rapid cluster dispersal ( kim et al .
2000 ; portegies - zwart et al .
2002 ) .
thus there are hints that the current generation of milky way starburst clusters is not long - lived , and hence is different from proto - globular clusters .
recently , stolte et al . ( 2008 ) compared multi - epoch high - resolution adaptive optics observations of arches , and derived an upper limit on the 2d velocity dispersion in agreement with the radial velocity measurements .
they also discuss that astrometric follow - up observation should yield the true velocity dispersion of arches .
ongoing multi - epoch astrometric monitoring of milky way starburst clusters will thus provide considerably improved constraints on the internal velocity dispersion , which in turn will be valuable for comparison with theoretical models .
the next generation of high - precision astrometric instruments like gravity for the eso vlti should reveal the kinematics of stars in the very cores of the starburst clusters , provide dynamical mass estimates for the most massive stars , and possibly even trace intermediate mass blackholes hidden in the very centers of these clusters ( gillessen et al .
2006 ) .
many thanks to boyke rochau and andrea stolte for comments and help with the figures .
this paper is based in part on data relase 3 + of the ukidss galactic plane survey ( warren et al .
2008 , in prep ) .
the ukidss project is defined in lawrence et al ( 2007 ) .
ukidss uses the ukirt wide field camera ( wfcam ; casali et al , 2007 ) and a photometric system described in hewett et al ( 2006 ) .
the pipeline processing and science archive are described in irwin et al ( 2008 , in prep ) and hambly et al ( 2008 ) .
adams , f. 2000 , apj 542 , 964 ascenso , j. et al .
2007 , a&a 466 , 137 baumgardt , h. , kroupa , p. 2007
, mnras 380 , 1589 brandner , w. , clarke , j.s . , stolte , a. , waters , r. , negueruela , i. , goodwin , s.p .
2008 , a&a 478 , 137 davies , b. , figer , d.f . , kudritzki , r .-
p . , mackenty , j. , najarro , f. , herrero , a. 2007 , apj 671 , 781 figer , d.f . ,
mclean , i.s . ,
morris , m. 1999 , apj 514 , 202 figer , d.f . ,
najarro , f. , gilmore , d. , morris , m. , kim , s.s . , et al .
2002 , apj 581 , 258 figer , d.f .
et al . , 2006 ,
apj 643 , 1166 geyer , m.p . ,
burkert , a. 2001 , mnras 323 , 988 gillessen , s. , perrin , g. , brandner , w. , straubmeier , c. , eisenhauer , f. et al .
2006 , spie 6268 , 33 hills , j.g . 1980 ,
apj 225 , 986 kim , s.s . , figer , d.f . ,
lee , h.m . , morris , m. 2000 , apj 545 , 301 lada , c. , j. , margulis , m. , dearborn , d. 1984 , apj 285 , 141 mengel , s. , tacconi - garman , l.e .
2007 , a&a 466 , 151 palla , f. , stahler , s.w .
1999 , apj 525 , 772 portegies zwart , s.f . ,
makino , j. , mcmillan , s.l.w . , hut , p. 2002
, apj 565 , 265 siess , l. et al .
1997 , a&a 326 , 1001 siess , l. , dufour , e. , forestini , m. 2000 , a&a 358 , 593 stolte , a. , brandner , w. , brandl , b. , zinnecker , h. , grebel , e. 2004 , aj 128 , 765 stolte , a. , brandner , w. , grebel , e. , lenzen , r. , lagrange , a .- m .
2005 , apj 628 , l113 stolte , a. , ghez , a. , morris , m. , lu , j.r . ,
brandner , w. , matthews , k. 2008 , apj , 675 , 1278 whitmore , b.c . ,
schweizer , f. 1995 , aj 109 , 960 zinnecker , h. , yorke , h.w .
2007 , araa 45 , 481 | because of their large number of stars spread over the entire stellar mass spectrum , starburst clusters are highly suitable to benchmark and calibrate star formation models and theories . among the handful of galactic starburst clusters ,
westerlund 1 with its estimated 150 o - stars , w - r stars , supergiants and hypergiants is the most massive young cluster identified to date in the milky way .
while previous studies of westerlund 1 focused largely on optical and x - ray observations of its evolved massive stellar population , we have analyzed near - infrared data , resulting in the first in depth study of the `` lower - mass '' main sequence and pre - main sequence cluster population , i.e. , of stars in the mass range 0.4 to 30 solar masses .
the derived properties of the cluster population allow us to test theoretical evolutionary tracks .
by comparison of westerlund 1 s half - mass radius with younger starburst clusters like ngc 3603yc and arches , and somewhat older massive clusters like rsgc1 and rsgc2 , we find evidence for a rapid dissolution of galactic starburst clusters , which has interesting implications for the long - term survival of starburst clusters , and the question to which extent galactic starburst clusters might mimic proto - globular clusters . |
we perform full quantum state tomography of all experimentally generated graph states , except the 7-qubit state @xmath89 as the number of measurements required for complete state tomography in this case is impractical .
nevertheless we still confirm the non - classical nature of the 7-qubit state via a multipartite bell inequality , described in the next section .
density matrices are reconstructed , from experimentally estimated expectation values , using the maximum likelihood method @xcite .
we measure an over - complete set of expectation values corresponding to all @xmath90 combinations of tensor products of the pauli operators .
the experimentally reconstructed density matrices , which are ideally states @xmath91 , @xmath92 , @xmath93 and @xmath94 , are presented in figures [ lc4 ] , [ ec1 ] , [ ec3 ] and [ rc4 ] , respectively . from these density matrices , properties such as the ideal - state fidelity and purity
are calculated ( see table 1 in the main text ) .
errors are determined via monte carlo ( mc ) simulation of projection noise ( due to a finite number of measurements ) centred around the experimentally estimated probabilities @xcite .
the mc simulations produce a distribution of ` noisy ' density matrices , from which uncertainties in derived properties can be estimated .
table 1 in the main text presents results from measurements of multipartite bell inequalities .
we were able to show that all our experimentally generated graph states violate such an inequality .
we now give more details on this .
let @xmath95 denote the stabilizing operators of an @xmath23-qubit graph state @xmath96 associated with the graph @xmath4 .
the group of their products , @xmath97 , is called the stabilizer @xcite .
it is given by @xmath98 with @xmath99 where @xmath100 denotes a subset of the vertices of @xmath4 .
the normalized bell operator is then defined as @xmath101 let @xmath102 where the maximum is taken over all local hidden variable ( lhv ) models .
any graph state @xmath96 fulfills @xmath103 .
consequently one obtains a bell inequality whenever @xmath104 .
it reads @xmath105 . as an example , the normalized bell operator for the graph @xmath106 is given by @xmath107 the value @xmath108 is given by @xmath109 @xcite . in the experiment
we find @xmath110 and thus a violation of the bell inequality by many standard deviations ( see table 1 in the main text ) .
the value of @xmath111 for the three and four qubit linear cluster as well as for the box cluster state can be found directly in @xcite .
a bound for @xmath112 , where @xmath113 is the graph underlying the five qubit state @xmath114 ( used to demonstrate quantum error correction ) , can be found in the following way .
first one notes that @xmath114 is equivalent to the state @xmath115 in figure [ bell5 ] b ) up to local clifford ( lc ) operations .
the two graph states have the same rank indices and are thus equivalent up to local unitary operations @xcite .
the fact that they are both graph states then implies the lc equivalence .
the local clifford operations do not change the value of @xmath112 . the graph state @xmath115 is built from a four qubit ghz state @xmath116 and a single qubit graph @xmath117 , connected by an edge
. application of lemma 3 in @xcite then gives a bound on @xmath112 : @xmath118 in a similar way one can bound the value @xmath119 , @xmath120 b ) @xmath115.,title="fig : " ] ( -190,70)a ) ( -70,70)b ) ( -90,32)@xmath121 ( -92,40)lc a straightforward calculation shows that the normalized bell operator equals the projector onto the graph state : @xmath122 consequently , one can determine the fidelity @xmath123 of an experimentally obtained graph state with density matrix @xmath124 by measuring the normalized bell operator . for all graph states , except the 7-qubit state @xmath125 , we have the state fidelity measured in two different ways : 1 . from the reconstructed density matrix and 2 . from @xmath126 .
the results are presented in the table in the main text .
both ways give the same results to within statistical uncertainty of one or two standard deviations .
one should not expect the values to be exactly the same since only a subset of the tomographic measurements are used in calculating @xmath126 , giving different statistical contributions .
we created the 4-qubit box cluster state @xmath127 , which is an example of a ring cluster @xcite and the smallest intrinsically 2d cluster state .
the state @xmath127 can be generated by preparing all qubits in @xmath14 and applying cphase gates between every pair of qubits connected by an edge . i.e. @xmath128 , where @xmath129 is a cphase between qubits @xmath130 and @xmath131 .
we experimentally generate a state @xmath94 which is equivalent to @xmath127 up to single - qubit unitary operations ( i.e. they are locally equivalent ) .
specifically , the local unitary operators , which have to be applied to @xmath94 to convert it to @xmath127 are shown in table [ tablecorbox ] . the laser pulse sequence used to generate
the state @xmath94 is the same as that to generate the error correction graph states , and is shown in figure [ gates_ec ] . | measurement - based quantum computation ( mbqc ) represents a powerful and flexible framework for quantum information processing , based on the notion of entangled quantum states as computational resources .
the most prominent application is the one - way quantum computer , with the cluster state as its universal resource .
here we demonstrate the principles of mbqc using deterministically generated graph states of up to 7 qubits , in a system of trapped atomic ions .
firstly we implement a universal set of operations for quantum computing .
secondly we demonstrate a family of measurement - based quantum error correction codes , and show their improved performance as the code length is increased .
we show that all our graph states violate a multipartite bell inequality and are therefore capable of information processing tasks that can not be described by a local hidden variable model .
the methods presented can directly be scaled up to generate graph states of several tens of qubits .
the circuit model of quantum computation is conceptually similar to a classical computer : a register of two - level systems in a simple initial product state are manipulated using unitary quantum logic gates @xcite .
mbqc @xcite represents a conceptually and practically different approach : after preparing an entangled cluster state of qubits @xcite , computation proceeds by performing measurements and feedforward . both approaches present different theoretical and practical challenges to realisation and warrant investigation in parallel .
recently , researchers have found novel applications for mbqc beyond universal qc , including e.g. blind quantum computation @xcite measurement - based entanglement purification @xcite and quantum error correction @xcite , featuring very high thresholds .
owing to the two - stage process of mbqc resource creation followed by its processing resources states can be purified and manipulated beforehand .
this offers a large degree of flexibility in optimizing and compressing schemes for quantum information processing .
schemes for correcting errors in universal mbqc have also been found with extremely high tolerence to errors , compared to those known for the circuit model of qc @xcite .
important experimental progress on mbqc has been made using entangled states of up to 8 photonic qubits @xcite . scaling up the non - deterministic methods used to generate entangled states in these works is very challenging , since their success probably reduces exponentially in photon number .
very recently there has been work on generating cluster states in continuous variables of light fields @xcite . in this work
we present the first demonstration of mbqc using trapped ions .
furthermore , we make two experimental steps forward in the model of mbqc that are system - independent : the deterministic generation of cluster states and the demonstration of quantum error correction ( qec ) .
the paper is organised as follows ; firstly mbqc is briefly reviewed and our approach to preparing cluster states is summarised ; then a universal set of operations is presented using a 4 qubit cluster state ; finally the phase - flip correction code is demonstrated for increasing codeword lengths .
we do not implement active feedforward , which has previously been demonstrated with trapped ions @xcite .
our results are post - processed to reproduce the action of perfect feedforward .
a mathematical graph @xmath0 is a set of vertices @xmath1 and edges @xmath2 .
the corresponding _ graph state _ is a physical state of @xmath3 qubits , associated with the vertices of the graph @xmath4 , which is defined in the following way . for every vertex one defines an operator @xmath5 where @xmath6 denotes the neighborhood of vertex @xmath7 and @xmath8 and @xmath9 denote pauli spin-@xmath10 operators .
the graph state @xmath11 is then uniquely defined as the common eigenstate of all operators @xmath12 with eigenvalue @xmath13 .
one approach to generating graph states is to start from an initial state with all qubits in @xmath14 , i.e. the + 1 eigenstate of @xmath8 , and then apply a controlled phase ( cp ) gate @xcite between every pair of qubits connected by an edge .
an important graph state is the 2d cluster state @xmath15 @xcite , which has the topology of a square lattice and belongs to the class of universal resource states @xcite .
any quantum computation can be carried out on a sufficiently large @xmath15 . in particular ,
any quantum logic circuit can be translated to a single - qubit - measurement pattern on @xmath15 @xcite .
measurements in the computational basis can be used to remove qubits from the cluster and imprint any desired quantum circuit structure onto the lattice .
measurements in the basis @xmath16 , where @xmath17 with real @xmath18 , drive the computation .
the value of @xmath18 determines , for example , the angle of single - qubit rotations . in general @xmath18
has to be adapted to the outcomes of previous measurements and thus introduces a temporal order and the need for feedforward @xcite .
while the use of the state @xmath15 offers a canonical route to quantum circuit simulation , the conceptual framework of mbqc is much more flexible and offers different and more economic modes of operation . instead of starting with @xmath15 , for example
, one can use graph states as resources which are tailored for specific quantum circuits , which could also be subroutines of a larger quantum algorithm . for circuits that contain mainly clifford gates ( including
e.g. the cnot and the hadamard gate ) their resource states can be highly compressed without loss of functionality .
such a situation arises for example in qec , which will be described below .
concatenating such tailored resource states in measurement - based qec and entanglement purification can give rise to much higher thresholds @xcite experiments use strings of @xmath19ca@xmath20 ions in a linear paul trap , manipulated with lasers .
two electronic states encode a qubit ( @xmath21 , @xmath22 ) that are coupled by an electric quadrupole transition at 729 nm .
we now briefly summarise how graph states are generated , for more details see the supplementary material .
experiments begin by preparing @xmath23 ionic - qubits in the product state @xmath24 via optical pumping , and preparing the axial centre - of - mass ( com ) and stretch ( str ) vibrational modes in the ground state by resolved sideband cooling .
graph states are generated using three distinct tools .
firstly an effective long - range qubit - qubit interaction of the form @xmath25 can be turned on for arbitrary times .
this interaction is realised by off - resonantly driving the axial com vibrational mode of the ion string using a bichromatic laser field from a single direction @xcite . when applied to ionic - qubits in @xmath24
this periodically generates fully connected graph states that are equivalent to ghz states @xcite .
the second and third tools both allow the selective removal of any connection in the graph @xcite . both use laser pulses tightly focused on individual ions . in combination these three tools allow , in principle ,
any cluster state to be deterministically created . in the circuit model of qc a universal set of logic gates
provides the tools to implement arbitrary quantum algorithms @xcite .
a common universal gate set consists of the cp gate and arbitrary single - qubit rotations around two independent axes , e.g. the z axis , @xmath26 and the x axis , @xmath27 , by angle @xmath18 @xcite .
all of these gates can be translated to carrying out specific sequences of measurements and feedforward on a four qubit linear cluster state @xmath28 , which are presented in figures 1a and 2a @xcite .
realising measurement patterns like these on large - scale cluster states , when combined with qec , enables arbitrary mbqc @xcite .
we create @xmath28 using a laser pulse sequence lasting 300@xmath29s and reconstruct the full density matrix via quantum state tomography ( see supplementary material for details ) .
the observed fidelity with the ideal state is @xmath30 , which is well above the threshold for multipartite entanglement of 0.5 @xcite .
measurement of @xmath28 in the order presented in figure 1 is equivalent to a circuit performing a sequence of one - qubit gates on the encoded state @xmath14 .
the choice of measurement basis of qubits 1,2 and 3 [ @xmath31 , @xmath32 and @xmath33 determines the overall rotation applied to @xmath14 .
we implement a range of measurement combinations , each demonstrating a different one - qubit rotation .
one approach , which avoids the need for active feedforward , is to reconstruct the output state ( encoded in qubit @xmath34 ) post - selected on the cases where the @xmath13 outcomes of the measurement of qubits 1 , 2 and 3 are observed , as in @xcite .
more information is obtained if all outcomes are kept and post - processed to simulate perfect feedforward .
results obtained in this way provide an upper limit for the performance that could have been achieved using feedforward .
figure 1 presents the results on the bloch sphere : a range of different rotated output states , reconstructed via quantum state tomography .
the average output state fidelity with the ideal case is @xmath35 .
measurement of @xmath28 in the order presented in figure 2 is equivalent to a circuit composed of a cp gate and one - qubit rotations , which operates on the encoded state @xmath36 .
the choice of measurement basis of qubits 1 and 4 determine the one - qubit rotations .
the output state is stored in qubits 2 and 3 .
we choose two important cases : a maximally entangled state ( case 1 ) and a product state ( case 2 ) are ideally created .
we quantify the generated entanglement by the tangle @xmath37 @xcite . in case 1
we find that the experimentally reconstructed state is strongly entangled , @xmath38 , and has a fidelity of @xmath39 with the ideal state . in case 2
the experimental state is close to being separable , @xmath40 , and has a fidelity of @xmath41 with the ideal state .
experimentally reconstructed two - qubit output density matrices are presented in figure 2 . taken together ,
the results in figures 1 and 2 demonstrate a universal set of operations .
we also created the 4-qubit box cluster state @xmath42 , an example of a ring cluster @xcite and the smallest intrinsically 2d cluster state .
we tomographically reconstructed the full density matrix and observed a fidelity of @xmath43 with the ideal state . for more details
see supplementary material . in a realistic setup one can not decouple the qubits on which the computation is performed completely from the environment , which will introduce errors on the qubits .
qec codes @xcite provide a solution by encoding the states @xmath44 and @xmath45 of a qubit into the states of larger physical systems @xmath46 and @xmath47 ( called codewords or logical qubits ) and using correlations to protect the information .
qec consists of at least two steps after the encoding .
firstly , one measures the correlation operators which reveal the error syndrome . in this step errors , which might be unitary qubit rotations or involve entangling to the environment ,
are discretized .
the discretization of the errors is a crucial step , as it reduces the infinite set of possible quantum errors to a finite set .
secondly , one applies the recovery operator to undo the error .
the principles of qec in the circuit model have been demonstrated before @xcite , including the 3-qubit phase - flip code @xcite .
circuit model qec codes can be translated to mbqc in the same way as algorithms .
however , it is important to note that qec codes involve only clifford gates and pauli measurements , which can be implemented in a very compact way in mbqc @xcite .
we demonstrate an mbqc phase - flip code , with codewords @xmath48 and @xmath49 , capable of correcting full phase flips ( z ) on up to @xmath50 of the codeword qubits . the general form of the graph state employed , labelled @xmath51 , and the protocol are presented in figure 3 .
@xmath51 consists of @xmath52 qubits : @xmath23 for the codeword , labelled @xmath53 to @xmath54 , and two additional qubits @xmath55 and @xmath56 to read in ( encode ) and read out the initial and final protected 1-qubit state , respectively .
for arbitrary @xmath23 , @xmath57 after preparation of @xmath51 , the protocol proceeds as follows : 1 .
a 1-qubit state @xmath58 or the orthogonal state is encoded by measuring qubit a in a basis where these are the eigenstates .
the effect is to distribute either state non - locally amongst the remaining @xmath59-qubits .
2 .
each of the @xmath23 central qubits @xmath54 is measured in the x basis , yielding one of @xmath60 possible outcomes .
this simultaneously decodes the state and reveals which of up to @xmath50 errors have occurred on the central qubits ( i.e. determines the error syndrome ) .
3 .
a 1-qubit correction operation , determined by the outcome in 2 is applied to the output state , stored in qubit @xmath56 , recovering the encoded 1-qubit state .
the temporal order of the measurements is unimportant and errors can happen to the central qubits , @xmath53 to @xmath54 , at any time before measurement of these central qubits .
it is useful to interpret the protocol as attempting to teleport a state across the cluster , from a to b , through a noisy channel affecting the middle qubits .
we demonstrate the protocol using the @xmath611 , 3 and 5 cases shown in figure 4.a - c .
equivalent experimental investigations of increasing codeword lengths in the circuit model have not yet been realised , due to the complexity of the gate sequences required .
the laser - pulse sequences used to generate each graph are described in the supplementary material . for the @xmath62 and 3 cases
we reconstruct the full @xmath63-qubit density matrices via quantum state tomography , yielding state fidelities of @xmath64 and @xmath65 , respectively .
the codes are tested against errors realised by applying 1-qubit rotations @xmath66 to all or a subset of the codeword qubits @xmath54 .
after measurement of @xmath54 ( and therefore discretisation of the errors ) this is equivalent to independent phase flips occurring incoherently and independently on those qubits to which it is applied , with probability @xmath67 . for input states we choose to encode the four eigenstates of the pauli x and y operators , which are maximally affected by phase flip errors .
error correction performance is quantified by the average teleportation state fidelity ( atf ) , through the noisy channel , averaged over the four input states .
firstly each code is tested against errors applied to one codeword qubit ( @xmath53 ) .
the @xmath61 3 and 5 graphs should be robust to this , whilst the atf for the @xmath61 1 graph should reduce linearly with @xmath68 , since it provides no error correction .
the results , presented in figure 4 , show quantitative agreement with the ideal cases , up to deviations that largely correspond to an overall fidelity drop due to imperfections in the graph state preparation .
also shown is the resistance against two errors ( @xmath53 & @xmath69 ) for the @xmath70 case , afforded by the increased codeword length .
we emphasise the quality of the results : even in the presence of large amount of noise we are able to teleport states across a 7 ionic - qubit string with fidelities of over 0.8 .
these diagnostic tests show that the experimentally generated graph states respond correctly to errors applied to individual codeword qubits .
a more realistic situation is that _ all _ codeword qubits are subject to error with the same probability .
figure 5d shows the theoretical performance of the ideal graphs against such noise : for @xmath71 , graphs with larger @xmath23 perform better , tending towards perfect correction up to @xmath72 as @xmath73 .
this is a challenging experiment since the small improvement offered by the larger ideal graphs is mitigated by the reduced quality with which the larger graphs were generated .
errors are applied to all physical codeword qubits in the experimentally generated graph states and the results are presented in figure 5e .
qualitative agreement with the ideal case is observed . even though many more qubits are exposed to errors in the larger codewords ,
there is still a region where they perform better .
that is we are able to demonstrate that , for a range of noise levels , a better protection of quantum information is provided when using a larger error correction code . for more discussion see supplementary material .
table [ stats ] summarises important properties of all experimentally generated graph states .
as shown in @xcite , any ideal graph state violates a multipartite bell - type inequality , i.e. it can not be described by a local hidden variable model . by measuring the expectation value of the multipartite bell observable @xmath74 , described in the supplementary material , we find a clear violation of the corresponding inequality for all prepared graph states .
these results imply that the information processing shown in the experiment is intrinsically quantum , and can not be described by a local hidden variable model .
furthermore , @xmath74 is equal to the state fidelity , but requires only a small fraction of the number of measurements for state tomography .
our largest error sources , which limit graph state quality , are laser intensity fluctuations and random electric fields that heat the ion string . using ionic - qubits we have made several distinct steps forward in mbqc : the deterministic generation of graph states , together with their application as resources ; the demonstration of measurement - based quantum error correction and ; the observation of improved performance with increasing codeword length
. it should be possible to directly scale - up the presented techniques to generate cluster states of several tens of qubits in the next few years .
more efficient schemes for generating ionic cluster states might employ direct nearest - neighbour interactions , such as those afforded by multi - site ion traps @xcite . both the circuit and measurement - based models of qc
have now been demonstrated in trapped ions . since
all instances have been proof - of - principle experiments , there is as yet no obvious reason to choose one over the other at this stage . both paths present similar challenges in terms of improvements over our ability to control many - body quantum systems .
+ + bpl acknowledges support by a marie curie fellowship ( piif - ga-2010 - 275477 ) .
this work was supported by the austrian science fund ( fwf ) under grant numbers : p25354-n20 , p24273-n16 and sfb f40-foqus f4012-n16 .
38 natexlab#1#1bibnamefont # 1#1bibfnamefont # 1#1citenamefont # 1#1url # 1`#1`urlprefix[2]#2 [ 2][]#2 , _ _ ( , ) , isbn . , * * , ( ) . , * * , ( ) .
, , , pp . ( ) . , , , , , , * * , ( ) . , , , * * , ( ) . , , , ( )
. , * * , ( ) . , * * , ( ) . , , , , , , , , * * , ( ) . , , , , , , , * * , ( ) . , , , , , , , , , , , * * , ( ) . , , , , , , , , , , ( ) . , , , , , , , , , , , * * , ( ) . , , , , , , , , * * , ( ) . , , , , * * , ( ) . , , , , * * , ( ) .
, , , , , , * * , ( ) . ,
* * , ( ) . , , , , , , , , , , * * , ( ) . , * * , ( ) . , , , * * , ( ) . , * * , ( ) . , , , , , , , , , , , * * , ( ) . , , , * * , ( ) . , , , , * * , ( ) . ,
* * , ( ) . , * * , ( ) . , , , , , , , , * * , ( ) . , , , , , , , , , , ,
* * , ( ) . , , , * * , ( ) .
, , , , , , , * * , ( ) . , , , * * , ( ) .
, , , , , , , , * * , ( ) . , , , , , * * , ( ) . , , , , , ,
* * , ( ) . , , , * * , ( ) . , , , ,
* * , ( ) .
. [ table ] * properties of experimentally generated graph states*. fidelity and purity derived from the tomographically reconstructed density state ( @xmath75 ) .
@xmath76}$ ] , @xmath77}$ ] , where @xmath58 is the ideal state .
@xmath78@xmath79 is equivalent to the state fidelity , derived from a subset of tomographic measurements .
values on the rhs of the inequality are the maximum allowed by lhv models ( see supplementary material ) .
nm : not measured .
errors are one standard deviation and derived from quantum projection noise . [ cols="^,^,^,^,^",options="header " , ] figures 4 and 5 in the main text present results for our measurement - based quantum error correction demonstration . in both cases the performance is quantified by the average teleportation fidelity ( atf ) , where the average is taken over the four input states @xmath80 , @xmath81 , @xmath82 and @xmath83 .
figures [ all6_local ] and [ all6_global ] in this text present results , from the same experiments shown in figures 4 and 5 in the main text , respectively , but now the atf is taken over six input states .
these six are the previous four plus the eigenstates of the pauli @xmath9 operator : @xmath44 and @xmath45 .
ideally these two new states should be perfectly teleported across the graph - states , independent of the error probability @xmath68 , since they are eigenstates of @xmath9 errors .
experimentally we find that these states have an imperfect fidelity due to the imperfect preparation of the graph - states .
these additional states do not probe the error correction capability of the graphs .
the conclusion that we draw from figure [ all6_local ] is the same as in the main text : the graphs respond correctly to errors applied to specific codeword qubits ( @xmath54 ) and the larger graphs are able to cope with errors on multiple qubits .
the conclusion that we draw from figure [ all6_global ] is also unchanged : there is a region where the larger codewords perform better , in spite of the increased complexity of their generation .
the size of this region is smaller when including all 6 states .
figures 4 and 5 in the the main text show deviations from the ideal cases .
we looked into this is in detail and now report on two observations , which are described below .
firstly , since full density matrices of the @xmath62 and @xmath84 graph states were measured , these can be used to simulate the outcome of the error correction tests that we performed .
specifically the entire error correction protocol , including the addition of errors , can be straightforwardly carried out by appropriately rotating and projecting the reconstructed density matrices .
the results calculated in this way are presented in figure [ err_rho ] c and compared to the results that are presented in the main text .
the different approaches show a close agreement . from this
we can conclude that the main sources of deviation between experiment and ideal theory are due to the errors in the experimentally generated graph states themselves , rather than mistakes in the way we precisely apply errors .
the high fidelity of the reconstructed states reflects the small errors they contain .
error budgets for two ionic - qubit logic gates have been given before @xcite .
specific and detailed experiments will have to be design to determine the extent of potential noise sources in multi - qubit operations .
however , it is clear that contributing error sources are laser intensity fluctuations and random electric fields that lead to am incoherent heating rate of the ion string common vibrational modes .
we observe a heating rate of approximately 50 phonons per second , of the axial com at 1.2 mhz . for @xmath61 1 and 3 , respectively .
rotations ( errors ) @xmath85 are applied to all central qubits ( in blue ) , which after measurement are projected into independent z flip errors with probability @xmath86 .
* c * performance calculated from experimentally reconstructed density matrices of the state in a ) solid blue line and b ) solid red line . *
d * experimental results by directly measuring the states in a ) blue diamonds and b ) red squares directly . ]
secondly , it was surprising to us that the average teleportation fidelity for the @xmath87 state , at @xmath88 , is not significantly lower than for the smaller states : this larger 7-qubit state is certainly experimentally more challenging to generate .
we looked into the effect of adding various types of noise to the ideal graph states on the teleportation fidelity at @xmath88 ( i.e. without additional @xmath9 errors ) .
we found that the larger states are more robust to various types of noise , such as depolarisation .
this may be understood by considering that any component of noise that leads to @xmath9 errors is better protected by the larger states .
indeed , one should expect that any part of the experimental noise , introduced in our graph - state generation process , that is equivalent to @xmath9 errors on the ideal state is better corrected by the error correction protocol for the larger states . |
the discovery of the acceleration of the expansion of the universe @xcite has revived questions concerning the domain of validity of einstein s general relativity . at present , the data gathered on the expansion rate do not disclose whether the acceleration is due to a component formally equivalent to the cosmological constant introduced by einstein @xcite , or whether we are finding the effective behavior of a theory of a wider scope whose low energy limit slightly departs from gr .
the bare einstein s equations ( without cosmological constant ) for a universe with a frw metric and dust like matter imply a continuous deceleration of the expansion rate .
however , a universe containing a fluid with an equation of state @xmath1 with index @xmath2 overcomes the deceleration when the density of this fluid dominates over that of the dust like matter . in this context , the cosmological constant , if it is positive and added to the equations , balances the deceleration by acting as a fluid with an equation of state @xmath3 .
the cosmological constant was originally in the metric part of einstein s equations , and alternative metric theories of gravity provide a wealth of terms in that side .
when describing a homogeneous and isotropic universe within a frw model , it is frequent to come as close as possible to the standard friedmann
einstein equation for the expansion rate by bringing terms to the right hand side within the so called `` effective index '' @xmath4 of the cosmic fluid whose equation of state is @xmath5 . in the field of gravitation ,
ideas that appeared in the classical domain several decades ago , are now being brought back within the different context of quantum gravity theories .
scalar tensor theories of gravitation are being reconsidered .
scalar fields coupled to gravity were proposed as a way to incorporate mach inertial induction .
they gave rise to the brans dicke theory of gravitation @xcite , which is well restricted by solar system experiments@xcite .
superstring theory and m theory introduce a number of quantum scalar fields , the dilaton , the radion , the moduli @xcite .
they can play a role in the late acceleration of the universe . on the other hand , brane worlds inspired by m
theory provide modifications to gravity and new equations of cosmic expansion @xcite . in the brane world context
, gravity lives in the bulk of 3 + 1+d dimensions while the rest of the interactions are confined to the 3 + 1 brane .
some braneworld proposals do not explain the observed cosmic expansion .
however , it was realized @xcite that it is possible to obtain late time acceleration with one extra dimension .
this possibility is interesting as it accounts for the observed cosmic expansion without the need of a cosmological constant term .
acceleration results from the induced gravity term in the brane . the number and size of additional dimensions has been explored in a wide range of proposals .
the existence of additional dimensions and their size are constrained by tests of the gravity law @xcite . at the time when gravity theories of the scalar tensor type and others were examined ,
the solar system tests were the most restrictive ones .
nowadays , the theories have to confront the cosmological observations .
the theories have to allow big bang nucleosynthesis @xcite .
they have to fullfil the constraints from the growth of density perturbations and give correct predictions for the evolution of the growth factor
. such growth function of density perturbations requires careful evaluation for several modified gravity scenarios @xcite .
ultimately , gravity theories have to be consistent with the cmb data and with the expansion rate of the universe provided by supernovae .
supernovae enable to measure the rate of the expansion of the universe along redshift z. they do so by tracing the luminosity distance along z for a standard candle . if the universe is flat , as inferred from wmap @xcite , the @xmath6 relation is simply : @xmath7 while in the general case , @xmath8 where @xmath9 @xmath10 @xmath11 , @xmath12 , or @xmath13 for closed , flat , and open models respectively , where the curvature parameter @xmath14 , is defined as @xmath15 .
@xmath16 is the hubble parameter . for a frw universe containing matter and a general form of matter
energy @xmath17 with index of the equation of state @xmath18 , @xmath16 is given by : @xmath19 where @xmath20 is the present value of the hubble parameter and @xmath21 where k accounts for the geometry and @xmath22 is the present value of the scale factor .
@xmath23 @xmath24 is the density parameter for dark energy .
when @xmath25 varies along z , @xmath16 can be expressed as : @xmath26\right\}\end{aligned}\ ] ] one sees from ( 12 ) , that @xmath6 contains @xmath25 through a double integral .
this leads to limitations in recovering @xmath25 if @xmath6 data along z are scarce and with large errorbars .
data are often given as @xmath27 which relates to @xmath6(z ) as : @xmath28\end{aligned}\ ] ] the quantity within brackets is @xmath29 , the `` hubble - constant - free '' peak absolute magnitude of a supernova .
the data are given sometimes in the form of distance moduli : @xmath30 the measurement of @xmath25 through @xmath31 is the standard approach , though possibilities of measuring directly @xmath16 have been investigated ( for a review see @xcite ) .
this hubble law in ( 3 ) is derived from the einstein - hilbert action .
we will review other possibilities related to expansion histories arising from modified gravity theories and effective theories which contain scalar fields accelerating the cosmic expansion . in the present overview we concentrate on the empirical evaluation of those theories .
some of the models to be explored here might contain a ghost problem or have a subclass amongst them that present quantum instabilities .
we refer for this topic to recent updates in @xcite .
the idea of allowing gravity to live in extra dimensions arose to explain the hierarchy problem and constitutes one of the major challenges to the classical picture of a 4dimensional description of gravity @xcite .
the planck scale of 10@xmath32 gev , relevant for gravitation , is much larger than the electroweak scale , of about 1 tev of the standard model of particle physics .
according to the proposal by arkani - hamed , dimopoulos and dvali @xcite , the planck scale is not a fundamental scale , but its enormity is simply a consequence of the large size of new extra dimensions .
while gravitons can propagate in the new extra dimensions , the standard model fields are confined to our submanifold of 3 @xmath33 1 dimensions , the brane , which is embedded in a higher dimensional bulk . in this scenario ,
the planck scale in the bulk can be much smaller than in the @xmath34dimensional brane as @xmath35 the existence of extra dimensions in which gravity propagates produces a transition in the gravitation law at the compactification radius , proposed to be of the submillimetre scale . in the cosmological context
, this idea has some interesting consequences since it predicts a departure from gr .
as soon as the cosmological model was implemented for @xmath36 extra dimensions @xcite , it was noted that the brane cosmology leads to friedmann like equations very different from the standard ones @xcite . in the 5dimensional model by dvali , gabadadze and porrati ( dgp )
whose cosmology is explored by deffayet and collaborators in @xcite , the competition between the five dimensional term of the einstein
hilbert action and the four dimensional term induced in the brane leads to a modified friedmann equation which simulates an acceleration term without the need to include a cosmological constant .
the transition from the gravity law @xmath37 to the modified one is characterized by a scale @xmath38 , the so called cross over scale defined as @xcite : @xmath39 at large distances , the five dimensional term takes over and gravity gets weaker @xcite . if the parameter @xmath38 is chosen to be of the order of h@xmath40@xmath41 @xmath42 10@xmath43 mm , this is a choice for @xmath44 @xmath42 10100 mev . at short distances @xmath45 , newton s law
is modified by a logarithmic repulsion . at large distances
@xmath46 , the modification involves a factor @xmath47 .
the hubble expansion law shows significant departures from the standard friedmann expansion law : @xmath48 where the sum over @xmath49 stands for a sum over each component with equation of state @xmath50 , @xmath51 being the density parameter : @xmath52 and @xmath53 in the matter dominated epoch , the expansion equation is : @xmath54 in the original 5d model in @xcite , @xmath55 was estimated to be @xmath42 1 tev , which corresponds to a cross over scale @xmath38 of 10@xmath56 cm .
such @xmath38 led to incompatibilities with the solar system data .
when @xmath55 is taken to be @xmath57 1 gev , i.e. @xmath58 10@xmath59 cm , the conflict disappears .
high z supernovae give useful constraints to the parameter @xmath60 in ( 10 ) @xcite . in @xcite ,
two samples of high z supernovae are used in conjunction with measurements from baryon oscillations @xcite and the cmb shift parameter @xcite to estimate the favored values for @xmath60 and @xmath61 . @xmath60
@xmath10 0.125 for the supernova data in the riess sample from 2004 @xcite and @xmath60 @xmath10 0.13 for the snls data @xcite .
it is found in @xcite that , for a flat universe , this 5d modified gravity model gives a worse fit than gr plus cosmological constant .
further supernova samples would reexamine the empirical compatibility of the dgp model and other modified gravity models involving more dimensions .
those models can be explored through a phenomenological description .
the hubble law arising from gravity with extra dimensions can be expressed by adding @xmath62 terms to the standard friedmann equation @xcite : @xmath63 where @xmath38 is the cross over radius , i.e. the transition radius from a regime to another in the gravity law , and @xmath49 is the parameter to be empirically determined . from redshift z@xmath10 0 to z@xmath64 , the above friedmann equation gives an effective index of the equation of state @xcite : @xmath65 such effective equation of state evolves too fast to be compatible with present sne ia data , unless @xmath49 is small .
present data favor @xmath49 close to 0 @xcite .
though the most simple implementations lack empirical confirmation , room is left for several models .
cosmological sne ia are useful tests in restricting the cross over radius that relates to departures from gr .
a different kind of model in which the brane is embedded in a 5d bulk has been proposed by randall and sundrum @xcite .
contrary to the previous approach , here the bulk has a @xmath66 geometry as opposed to the flat geometry of the dgp model . in the original randall sundrum ( r s ) model , the brane has tension @xmath67 , i.e. cosmological constant , as well as the bulk @xmath68 .
the bulk cosmological constant is negative .
the resulting hubble expansion in the brane is the result of the compensating effect of both terms . to make zero the effective cosmological constant in the brane , @xmath69 , a fine tuning between the brane tension and the bulk cosmological constant is required .
the original proposal sets the effective cosmological constant , @xmath69 , to zero .
generally , a tiny deviation from exact compensation @xcite would lead to an effective cosmological constant at late times .
modifications of the original r s proposal include in the brane , in addition to tension ( cosmological constant ) , a matter energy density @xmath70 .
the total brane density @xmath71 is : @xmath72 this gives the modified friedmann equation @xcite : @xmath73 where @xmath74 is the inverse planck mass @xmath75 and @xmath76 is related to @xmath68 .
generalizations of the randall sundrum scenario in cosmology consider the _
ads@xmath77 _ bulk geometry containing the fifth dimensional cosmological constant @xmath68 , but explore arbitrary densities in the brane and in the bulk .
explorations of braneworld ideas that allow a late acceleration of the universe can be found in @xcite .
before the development of braneworld scenarios , dark energy as associated to a scalar field was considered in analogous way to research done within the context of inflation @xcite .
this theoretical possibility was entertained a decade before there was evidence of a non
zero cosmological constant from supernovae and that the cmb results indicated a flat universe .
the favored @xmath78 value by inflation and the evidence that @xmath61 @xmath79 1 led to think about a possible @xmath80 term . in the recent expansion history a scalar field could play a role in analogy with inflation , where a scalar field dubbed the inflaton gives rise to the exponential phase of early acceleration .
the scalar field responsible for the late acceleration , later named quintessence , is proposed as a new massless scalar field which induces a cosmological constant behavior at late times . in this
original setting , the field is minimally coupled to gravity , i.e. , thus the inclusion of the field remains within the framework of gr .
the interaction of the field with the matter fields is set to be negligible .
the scalar field responsible for dark energy can be identified by recovering its potential , or equivalently its effective equation of state .
the presence of such scalar field @xmath81 , would induce an action : @xmath82\ ] ] the field contributes to the stress
energy momentum tensor with an effective mass density and pressure : @xmath83 in this case , the effective value of the equation of state @xmath4 depends on the form of the potential @xmath84 and it can evolve with time @xcite .
@xmath85 a search for general properties of this potential , in analogy to the search for the inflation potential , leads to suggestions for potentials with tracking behavior @xcite .
several potentials have been proposed , from the original @xmath86 @xcite to other recent proposals @xcite .
the dark energy scalar field that does not couple with matter has the feature of avoiding effects on variations of constants and other effects subject to experimental evidence @xcite .
there is already a reconstruction of what could be the potential of the minimally coupled field @xmath84 obtained from the snls sne ia @xcite following reconstruction ideas given in @xcite .
another reconstruction of @xmath84 using sne ia from various collaborations and considering pad approximant expansions of the potential is found in @xcite . by now , many proposals have been discarded , since the equation of state between z@xmath87 and z@xmath88 evolves very slowly as shown by high
z sne ia data @xcite . one can still consider the case of a scalar field that is coupled to other matter fields or to gravity . some of these proposals come from the ballpark of superstring theory . the most well known scalar
tensor gravity proposal , i.e. brans dicke theory , incorporated the coupling of a field to gravity to account for the mach induction on newton s constant . in brans
dicke gravity @xcite , newton s constant is replaced by a dynamical field @xmath89 leading to the generalized friedmann equation : @xmath90 the brans dicke action contains a @xmath91 term : @xmath92\ ] ] the theory has been well tested @xcite .
observationally , for other scalar tensor theories , we aim at retrieving the self interaction potential of the scalar @xmath84 . and the coupling function to matter and gravity @xmath93 in the matter term of the action @xcite : @xmath94 in the case of quintessence ( minimally coupled field ) @xmath95 .
the present supernova data have explored the potential for a minimally coupled field @xcite . in the non - minimally coupled case , also referred to as coupled quintessence or coupled dark energy @xcite , the cmb data can put constraints on the strength of scalar gravity as compared to ordinary tensorial gravity @xcite . however , there are limitations to the recovery of the form of the coupling and the coupled dark energy potential from cmb observations @xcite .
supernovae together with observational constraints on the growth of density perturbations would allow to retrieve @xmath93 @xcite .
we expand @xmath93 in its derivatives @xmath96 , @xmath97 and higher orders : @xmath98 by exploring the empirical evidence it is found @xcite that solar system tests and binary pulsar tests allow to impose precise bounds on the first and second derivatives of the matter scalar coupling function , while sne ia could `` a priori '' allow to reconstruct the full shape of the function @xmath93 or the higher order derivatives .
it is argued in @xcite that the knowledge of the luminosity distance and the density fluctuations @xmath99 as functions of redshift z is sufficient to reconstruct the potential @xmath84 and the coupling function @xmath93 .
these results can be generalized to proposed actions that can be re expressed as a theory of interacting matter fields in general relativity @xcite .
actions containing @xmath100 are motivated from the lowest order effective action including the dilaton @xmath101 @xcite .
they can be brought by redefinition of the field to an einstein frame in which the field is minimally coupled to the metric . a number of interesting cases have been examined and have been expressed in terms of a scalar field which is minimally coupled to the metric , but nonminimally coupled to the other fields @xcite .
the self interaction potential shows an explicit energy transfer between the scalar field and the matter fields .
related to the previous exploration is that of _
f(r ) _ gravity or extended curvature gravity @xcite , also known as `` c essence '' when including an inverse power of r @xcite . those modified gravity theories with a lagrangian containing an arbitrary function of @xmath102 are equivalent to a particular class of scalar
tensor theories of gravity .
for instance , the action @xmath103 is found to be equivalent to @xcite : @xmath104 the equivalence is generalized to an arbitrary function of r @xcite : @xmath105 with an equivalent action @xcite : @xmath106 where @xmath107 .
the introduction of a canonical scalar field @xmath108 such that @xmath109 , allows to rewrite the action with f(r ) as : @xmath110 the gravity described by @xmath111 is the einstein
scalar system and @xmath112 ( @xmath113 ) .
@xmath112 is a metric of scalar tensor gravity which is , in this case , subject to constraints coming from solar system experiments .
again , a classification of when one recovers an action equivalent to the einstein hilbert case and when one recovers a brans dicke type action in _
f(r ) _ gravity is given in @xcite . when considering a general function of r , it is not always possible to obtain a description equivalent to einstein gr plus additional fields minimally coupled to the metric but coupled to other fields .
when this is possible , we have in the matter term the coupling between the matter fields @xmath114 and the scalar field @xmath81 .
examples of _
f(r ) _ casted in terms of fields can be found in @xcite . for the starobinsky model @xcite with @xmath115 with @xmath116 gev , for instance , the effective potential can be rewritten , in terms of the scalar field @xmath108 , as @xmath117 actions with generalized inverse powers of @xmath102 have been investigate as well @xcite .
the supernovae gathered up to now are compatible with some _
f(r ) _ gravity models and incompatible with others @xcite .
more general scalar
tensor theories than the ones here defined , such as those where the potential depends on several scalar fields @xmath118 and matter scalar coupling @xmath119 , give rise to a wider phenomenology . in the superstring scenario ,
nonminimal coupling of the scalar fields to the spacetime curvature leads to effective actions with similarities to scalar
tensor proposals .
some scalars present in the effective action arise from compactification to four dimensions .
those are structure moduli which do not couple directly with the spacetime curvature tensor .
one modulus field @xmath67 associated with the overall size of the internal compactificaton couples to einstein gravity via riemann curvature invariants , such as the gauss bonnet invariant @xcite . if one considers the gauss bonnet term , at least an extra function , the coupling to the gauss bonnet term , @xmath120 needs to be determined . in the following , we review attempts to include such enlarged action and their test with cosmological d@xmath0 data .
the gauss bonnet term can be found in the effective low energy string lagrangian and in brane theories @xcite .
the case of the simplest action containing the gauss - bonnet term can be written as : @xmath122\end{aligned}\ ] ] this lagrangian includes the gauss bonnet term coupled to a field @xmath123 , for instance the modulus field in @xcite where @xmath120 is the coupling of the modulus field to the gauss bonnet term , @xmath124 .
@xmath125 the possible coupling of a scalar field to the gauss - bonnet topological invariant can be constrained if one takes into account cosmological supernovae and solar system tests . from supernova data
gathered in @xcite and with a parameterized analysis of the evolution , it is found that the present value of the equation of state @xmath126 is below 1 .
though for the measurement of the index of the equation of state at @xmath127 one awaits a large enough nearby supernova sample , the results presented in @xcite move to examine the possibility of @xmath128 . in the original exploration of the idea @xcite , a universe which evolves towards @xmath4 @xmath79 1 would suffer a super aceleration phase , that has come to be called the big rip . from an action which includes the gauss bonnet term , one can obtain @xmath128 for a canonical scalar field ( positive kinetic energy ) @xcite .
it avoids the big rip for @xmath128 . to add a word on this
, this is not the main motivation for gauss bonnet gravity , since variations of the braneworld scenario can also get @xmath128 and avoid the big rip @xcite .
the interest on the gb term is justified by its presence in a variety of gravity proposals from string and m theory .
recently some tests using supernovae have already been done @xcite .
the extra gb term opens naturally many possibilities for the evolution of the scale factor .
the supernova test succeeds for a range of effective low energy actions including a gauss - bonnet term @xcite .
the fields that interact with the dark energy scalar field are constrained by a large number of tests @xcite .
the possibility that the dark energy scalar couples to nucleons and photons has been investigated in relation to implications for a fifth force @xcite . within current scenarios of particle physics such as mssm and msugra the role of this dark energy field shows some observational signatures of the fifth force type as well @xcite .
for instance , the interaction of the dark energy scalar field @xmath81 with the higgs doublet @xmath129 and @xmath130 leads to an action where the field couples differently to each of the higgs @xmath131 and @xmath132 and one expects a violation of the weak equivalence principle .
in the string scenario , where radion , dilaton and moduli can be responsible for the early and late acceleration , one has as well some possible strong couplings of the fields with matter @xcite .
thus , the cosmological context has to be considered together with classical gravitational tests on earth .
thus far , the torsion balance experiments set strict limits on the variation of newton s law using different substances and locations .
there is a wide range of tests being done to examine the predictions from scalar
tensor proposals and modified gravity ideas @xcite .
the empirical bounds on some types of couplings are tight and worth considering together with the cosmological probes .
finding that the negative pressure component gives @xmath133 , would likely be the biggest challenge to gravitation theories . at present , sne ia have not provided such evidence , though averaged values of @xmath134 are close to 1 at redshifts up to z@xmath42 1 . to probe that @xmath134 stays 1 up to the surface of last scattering ,
involves refining several other methods ( growth of density perturbations , integrated sachs
wolfe effect ) @xcite . from the view point of string theorists
@xcite , there is no natural explanation for the vanishing or extreme smallness of the vacuum energy .
a way to address the small value of the cosmological constant in the context of a stringy landscape is to consider the anthropic principle @xcite .
however , invoking the anthropic principle as a way out of the problem entails to abandon the quest for a conventional scientific explanation @xcite . .
from the observational point of view , the answer is near to @xmath135 , but possible departures from @xmath133 are presently being examined through expansions of the equation of state along the scale factor and other various approaches .
a expansion often used is @xcite : @xmath136 where the scale factor @xmath137 , so @xmath138 when evaluated at @xmath139 .
differences between the cosmological constant and departures from it in the equation of state are measurable nowadays ( see fig 1 ) . to reach a higher level of discrimination between dark energy models
depends on the progress made in the accuracy of the measurements .
the aim of future projects is to determine distances with 1 @xmath140 accuracy , i.e. with errors in magnitudes of 0.02 mag @xcite .
if the equation of state of dark energy is very close to @xmath141 , one can anticipate decades of improving the supernova measurements .
the development that we might witness could be akin to that of the postnewtonian tests of gr where higher levels of precision are being targeted since the middle of the last century into the present . ,
i.e. a model with @xmath142 and @xmath143 ( short dash line ) and a model with @xmath142 and @xmath144 ( dash - dotted line ) ]
a long path has brought us to the present situation where we can already aim at accuracies of 1@xmath140 in distance .
back in 1968 , a first attempt to draw the hubble diagram from sne resulted in a spread of @xmath67 @xmath42 0.6 mag @xcite .
such large scatter was due to the inclusion of other types of supernovae , which were later recognized as being of a different type ( sni b / c ) and characterized by a large scatter in luminosities @xcite .
sne ia are not standard candles either , but they follow a well known correlation between maximum brightness and rate of decline of the light curve . in 1977 , pskovskii @xcite first described the correlation between the brightness at maximum and the rate of decline of the light curve : the brighter sne ia have a slower decline of their light curves whereas fainter ones are the faster decliners . a systematic follow up of sne ia confirmed the brightness
decline relation @xcite . the intrinsic variation of sne ia
is written as a linear relationship : @xmath145 @xmath146 , the parameter of the sne ia light curve family , is the number of magnitudes of decline in 15 days after maximum . the value of @xmath49 , as well as the dispersion of that relation , have been evaluated in samples obtained from 1993 to the present @xcite .
@xmath147 is the absolute magnitude in @xmath148 for a template snia of @xmath149 rate of decline .
an intrinsic dispersion of 0.11 mag in that law is found when information in more than one color is available .
in fact , the 0.11 mag scatter in the brightness rate of decline is found when one includes supernovae affected by extinction . the scatter is smaller in environments where dust in the host galaxies of the sne ia is a lesser effect @xcite .
additionally , the family of sne ia form a sequence of highly resembling spectra with subtle changes in some spectral features correlated with the light curves shapes @xcite .
the multiple correlations allow to control errors within the family of sne ia @xcite .
the parameter @xmath146 requires to have observed the maximum in the light curve of the supernova .
this is not always possible , and therefore the various sn collaborations have formulated the brightness
decline correlation in different ways .
the high z sn team uses the full shape of the light curve with respect to a template , method refered as mlcs2k2 @xcite .
the supernova cosmology project collaboration , on the other hand , has introduced the _ stretch factor ,
s , _ as a parameter to account for the brightness
decline relationship @xcite .
the stretch parameter fits the snia light curve from premaximum to up to 60 days after maximum .
thus , even if the maximum brightness phase is not observed , the whole light curve provides the decline value .
more recently , other ways to measure the decline have been proposed @xcite .
measuring the evolution of the rate of expansion of the universe does not require knowledge of its present value , @xmath150 .
large projects have been devoted to the determination of @xmath150 by attempting the absolute calibration of @xmath151 from nearby sne ia that occur in galaxies with cepheid measured distances @xcite .
controversy persists , though the most recent values quoted by the two collaborations using this method are closer ( in the range 6575 km s@xmath41 mpc@xmath41 ) @xcite .
though largely ignored as an argument , the mere fact that a snia is a explosion of white dwarf bounds very efficiently the lower and upper possible values of @xmath150 @xcite . on the empirical side ,
the emission at late phases of sne ia gives a precise value of @xmath150 , since such emission is very sensitive to the density profile , the mass of the supernova and the amount of @xmath152ni synthesized in the explosion .
those quantities are easily measurable from the spectra @xcite .
the value favored by this method is h@xmath40 @xmath10 68 @xmath153 6 ( syst ) km s@xmath41 mpc@xmath41 .
this disgresion done , we emphasize the _ nuisance role _ of the absolute magnitude of sne ia when it comes to dark energy cosmology . the apparent magnitude of supernovae along z give us the cosmology , @xmath154 as @xcite : @xmath155 where @xmath156 is the stretch value for the supernova , @xmath157 is the `` hubble - constant - free '' luminosity distance , @xmath6 is given by ( 1 ) and @xmath158 is the `` hubble - constant - free '' @xmath159-band peak absolute magnitude of a @xmath160 sn ia with true absolute peak magnitude @xmath151 . similar use of the nuisance parameter @xmath161 in their light curve fitting is made in @xcite . in the first determinations of dark energy cosmology , there were four parameters in the fit : the matter density parameter @xmath61 , the cosmological constant density parameter @xmath162 , as well as the two nuisance parameters , @xmath163 and @xmath49 ( either @xmath49 itself as used by the scp or a equivalent parameter by other collaborations ) .
the four - dimensional ( @xmath61 , @xmath162 , @xmath163 , @xmath49 ) space , explored as a grid , yields @xmath164 and @xmath165 of the fit of the luminosity distance equation to the peak @xmath159band magnitudes and redshifts of the supernovae .
after normalization by integrating over the `` nuisance '' parameters , the confidence regions in the @xmath61@xmath162 plane can be obtained .
this use of the magnitude redshift relation @xmath27 as a function of @xmath61 and @xmath162 with a sample of high z sne ia of different z , is a very powerful pointer to the allowed values in the @xmath61@xmath162 plane @xcite . by 1998 ,
such use yielded important results , impling that @xmath166 at the 3@xmath67 confidence level . for a flat universe ( @xmath78 ) ,
the results from the supernova cosmology project meant @xmath61=@xmath167(statistical)@xmath168 ( systematics ) , and the high z supernova team obtained for a flat universe @[email protected] .
the outcoming picture of our universe is that about 2030@xmath140 of its density content is in matter and 7080% in cosmological constant .
according to the allowed @xmath61 and @xmath162 values , the universe will expand forever , accelerating its rate of expansion .
if the size of the sne ia sample and its redshift range do allow it , one can aim at determining the average value of the index of the equation @xmath169 or , going a step further , aim at determining the confidence regions on the @xmath126 @xmath170 plane .
for this later purpose , using the information on the global geometry as given by wmap provides a helpful prior . in knop
( 2003 ) @xcite , the supernova results were combined with measurements of @xmath61 from galaxy redshift distorsion data and from the measurement of the distance to the surface of last scattering from wmap .
the confidence regions provided by the measurement of the distance to the surface of last scattering and the sne ia confidence regions cross in the @xmath61@xmath134 plane , thus providing very good complementarity .
this work considers an averaged @xmath169 ( given the range in z of the sne ia observed ) . from the fitting assuming @xmath134 constant ,
the value is @xmath169 @xmath101.06@xmath171(statistical)@xmath172 ( identified systematics ) .
the analysis of the snls @xcite also aims at an average value of @xmath134 , allowed by the z range of this survey .
their 71 high z sneia give @[email protected] @xmath153 0.090 ( statistical ) @xmath153 0.054 ( identified systematics ) .
recently , the results from essence , using 60 sne ia centered at z@xmath10 0.5 point to a similar result with @xmath169@xmath10 1.05@xmath173 ( statistical ) @xmath153 0.13 ( systematics ) @xcite . in this last work ,
the supernovae from various collaborations are joined with two separate light curve fitting approaches giving a consistent result for the average value of @xmath134 .
high z sne ia beyond z @xmath174 1 enable to have a first look at possible evolution of @xmath25 .
the first campaign by riess et al .
( 2004 ) gave the results in the @xmath126@xmath175 plane , using the taylor expansion for @xmath25 as @xmath176 .
the best estimates for @xmath126 and @xmath175 are @xmath177 1.36@xmath178 and @xmath1791.48@xmath180 .
those results @xcite exclude a fast evolution in the equation of state , and therefore they rule out a number of modified gravity proposals .
moreover , the results are confirmed in @xcite where the available sne ia at large z , which can constrain @xmath126 and @xmath170 , are incorporated to test with a joined sample encompassing a wide range of z. significant limits have been placed on the evolution of @xmath25 .
now the potential of the method relies on an improved control of systematic effects . in this area ,
there has been substantial progress along the last years .
huge advances have taken place in controlling the uncertainties due to extinction by dust , the universality of the supernova properties , the control of lensing by dark matter and the observational process leading to the the construction of the hubble diagram of sne .
the study with the hst of the host galaxies of the sne ia sample of the supernova cosmology project @xcite allowed to subclassify the galaxy hosts of the p99 sample @xcite . it was found that , at high @xmath181 , early type galaxies show a narrower dispersion in sne ia properties than late type galaxies , as they do at low @xmath181 ( @xcite ) .
this result is very encouraging as it supersedes previous dispersion values obtained in samples of supernovae in all galaxy types .
supernovae in dust
rich environments , such as spiral galaxies , are more affected by extinction .
the supernova magnitudes are corrected regularly from galactic extinction @xcite , but the correction by dust residing in the host galaxy or along the line of sight is often not included in the standard fit , as it would require to have extensive color information that might not be available for high z sne ia .
when it is included , it is still subject to the uncertainty in the extinction law @xcite .
let us , for instance , consider the correction of dimming by dust as entering in the term @xmath182 , the magnitudes of extinction in the @xmath159 band .
we take , for instance , a supernova of a given stretch s , and we place its @xmath183 ( the magnitude of the equivalent supernova of @xmath160 ) in the hubble diagram .
we might use , as in @xcite : @xmath184 the extinction in the canonical band @xmath159 of the spectrum ( @xmath182 ) is determined by the extinction coefficient @xmath185 in that band . any uncertainty in this coefficient as due to variation of dust properties ,
leads to uncertain estimates of @xmath182 .
@xmath185 is of the order of 4.1 @xmath153 0.5 as properties of dust might change going to high z. if we have an error of 0.1 in the observable @xmath186 due to dust extinction , it gets amplified as : @xmath187 the present way of dealing with sne ia , including information in several photometric bands of the spectrum where extinction coefficients ( @xmath188 , @xmath189 ) are lower and one gets smaller values in the corresponding @xmath190 and @xmath191 , allows to control the dust problem .
in fact , the coefficients @xmath185 , @xmath188 and @xmath189 are also determined by methods determining distances to sne ia . in a sample of 133 nearby sne ia , it is found in @xcite , that the dust in their host galaxies is well described by a mean extinction law with @xmath188 @xmath192 2.7 . in all this , the level at which sne ia are treated , including how many bands are used , becomes critical .
a use of sne ia without taking into account color information leads to a dispersion of 0.17mag .
when taking into account color information , the limit goes down to 0.11mag .
limiting the effect of extinction , as in sneia in ellipticals should bring down the dispersion .
this finding is motivating the ongoing sn searches in clusters of elliptical galaxies @xcite . in environments where dust is controlled , such as sne in clusters rich in elliptical galaxies , the systematic effect introduced by dust extinction
can be better controlled .
there , one can also control the evolution of dust properties .
in table i , the level of accuracy in the first uses of sne ia back in 1998 is shown , with the contributions of each systematic and statistical error .
the improvement is given for a future mission able to control the systematics in the best possible way .
it is foreseen that several sources of systematic errors will become negligible within the next decade .
this question resides on whether we know that sne ia properties evolve with z. much has been done along this path .
nowadays , while we know that dust is the unavoidable systematics , several studies confirm the universality of the light curve properties of sne ia .
the study of a significant sample of sne ia at various @xmath181 by the supernova cosmology project reveals that the rise times to maximum of high z sne ia are similar to the low@xmath181 sne ia ( @xcite ) .
the same is found in the snls sample ( @xcite ) .
statistical evaluation is so far consistent with no difference between the low@xmath181 and high@xmath181 samples .
moreover , the possibility of existence of an extra parameter in the maximum brightness rate of decline relation has been carefully examined . among possible influences ,
metallicity was one of the obvious ones to consider . examining supernovae in galaxies with a gradient of metallicity , it is found @xcite that there is no evidence for metallicity dependence as an extra parameter in the light curve correlation of sne ia .
this result comes from an analysis of the sne ia light curves of a sample of 62 supernovae in the local universe .
the sne ia belong to different populations along a metallicity gradient .
this check has been done as well at high z @xcite using 74 sne ia ( 0.17 @xmath79 z @xmath79 0.86 ) from the scp sample .
no significant correlation between peak snia luminosity and metallicity is found .
often we address the question of the spread of the sne ia samples in galaxies of different types : on how many fast sne ia versus intermediate or slow decliners are found in the various morphological types .
spirals at low or moderate redshifts should encompass all ranges of variation in the sne ia properties since they contain populations with a wide spread in age .
the cosmological sn collaborations find that in those samples the one
parameter correlation does give a good description of the variation of the sne ia light curves : there is no residual correlation after the light curve shape correction @xcite .
one can go a step further and test the snia objects themselves .
this has been attempted in a intermediate z sample and it is a feasible project at all z @xcite .
the physical diagrams of sne ia at intermediate z and at low z seem to be similar @xcite .
those diagrams reveal the composition , velocity gradient and radiation properties of sneia @xcite .
looking at the inside of the sne ia is the best way to test whether they are the same at all z. such step forward is giving definitive results which reaffirm the validity of the sne ia method .
the progress done in this domain is summarized in table i. there is another element yet to be introduced here : it deals with the k correction .
correction is necessary to calculate the apparent magnitude in a @xmath193 filter band of a source observed in a @xmath194 filter . as we are comparing the observed flux of supernovae in a given spectral band along the expansion history , we need to take into account the effect that the photons of the supernova have shifted to redder wavelengths and spread over a different spectral wavelength range and the observed flux is different from that at emission .
@xmath195 where @xmath196s_{y}(\lambda ) d\lambda}}\right\ } + { \cal z}_{y } - { \cal z}_{x}\ ] ] where @xmath197 is the source spectral energy distribution ( a sn in this case ) , and @xmath198 is the transmission of the filter x. the @xmath199 term accounts for the different zero points of the filters @xcite . at present , this term amounts to an uncertainty of 0.02mag in the overall budget of sneia as distance indicators , but it can be decreased to @xmath200 mag .
the comparisons of supernovae with the fine time sequences of spectra along all phases available , are turning the uncertainty in the k
correction into a decreasingly small figure .
the large amount of spectral data on sne ia , and the good correlation with photometric properties , can indeed make this uncertainty very small .
gravitational lensing causes dispersion in the hubble diagram for high redshift sources .
there was an early concern on the systematics and bias that this effect could cause in the determination of the cosmological parameters using sneia @xcite .
it has been shown through the study of lensing in the highest z sne ia sample @xcite that the magnification distribution of sneia matches very well the expectations for an unbiased sample , i.e. , their mean magnification factor is consistent with unity
. the effect can be very well controlled by studying the galaxy fields @xcite . moreover ,
sne ia should show a negligigle systematics caused by gravitational lensing when having a few sneia per redshift bin @xcite .
this means that in future missions that effect should not have impact in limiting the accuracy of @xmath25 .
the better we deal with the light curve and spectral information of sne ia , the lower becomes the dispersion in the hubble diagram . if sne ia have a systematic error in their use ( independently from evolution ) , this is likely well below the accuracy to which we would like to determine distances ( 1@xmath140 ) .
we have not yet hit that brick wall .
using multi
epoch spectral information and accurate light curves in various filters should enable to test the power of these candles and their limiting irreduceable error @xcite .
supernovae collected along years show the effect of observations treated with various degrees of accuracy .
the dispersion around the best fit reflects that fact .
for instance , some samples would have information on only two colors and at very few epochs , making it difficult to control reddening , or supernovae had no spectrum confirming that they were sne ia .
the new samples will naturally have a higher degree of accuracy in magnitude and extinction for each supernova than previous samples .
the ones with poorer information have been placed in @xcite into the silver category ( as opposed to `` gold '' category ) . however , even within the gold sample the requirements have changed : the latest definition of gold sample demands higher quality standards than in @xcite . in some of these sne ia lists , distance moduli of supernovae are assembled but they have been obtained through procedures which deal differently with dust extinction .
the hubble diagram resulting from those compilations would naturally show a large dispersion . in @xcite , it is argued that a few supernovae of an early high z sn search sample are pulling @xmath25 towards evolution . the exploration done with a method independent from that in @xcite showed that the central value in the 1@xmath67 confidence varied according to the light curve and reddening fitter of those sne ia @xcite .
fortunately , supernovae enable to compare results obtained with different light curve fitting approaches .
the published photometry by the various collaborations can be used to bring together the sne ia by using a consistent procedure into the hubble diagram .
this was done for the sne ia considering available up to 2005 in @xcite .
this is done in @xcite for samples including the latest data from the higher z collaboration .
in most of the above discussion on dark energy , the frw metric is taken as the right metric for our universe .
modern cosmology supports that the universe is highly isotropic on average on high scales .
the main evidence is coming from the statistical isotropy of the cosmic microwave background radiation .
there are also cosmological observations supporting a matter distribution homogeneous on large scales , of the order of 100 h@xmath41 mpc .
given the observed structures of matter in voids and filaments , it is worth considering whether departures from the homogeneous and isotropic universe could alter the present discussion , in particular , the possibility that there were no need for dark energy , but instead , that the observations are the result of an effect of inhomogeneity @xcite . as pointed out in @xcite
, large samples of supernovae will soon allow to test the luminosity distance relation in various directions and redshift bins .
one can consider the luminosity distance d@xmath0 as a function of direction @xmath201 and redshift z @xcite .
the direction averaged luminosity distance is : @xmath202 this is the usual value retrieved by the sn collaborations .
but its directional dependence @xmath203 can be a test of the validity of the isotropy assumption .
the authors in @xcite propose to expand the luminosity distance in terms of spherical harmonics , to obtain the observable multipoles @xmath204 .
this idea was also investigated in @xcite , where the motion of our local group with respect to the cmb was measured .
results in @xcite are in agreement with the results in @xcite .
moreover , we can think of using the dispersion in the magnitude of sne ia in each direction @xmath201 to put constrains on anisotropic models . as discussed in @xcite ,
the dispersion of the magnitude in the hubble diagram is at odds with some simple inhomogeneous models that predict a higher spread in @xmath27 and also a non linear behavior .
the anisotropic and inhomogeneous models have to confront present and future values on the sneia dispersion at each redshift interval @xmath205 and in every direction @xmath201 : @xmath206 the scaling solutions of the regionally averaged cosmologies can simulate the presence of a quintessence field or other negative pressure component @xcite .
but , any backreaction from inhomogeneity and anisotropy would show in the dispersion in the hubble diagram .
those challenges have to be met by the non frw proposals . up to now ,
the exploration of toy models calculating the effect of inhomogenities in the scale factor , find irregularities in the expansion rate beyond what is observed in the sn ia hubble diagram .
the dispersion in the hubble diagram of a standard candle is increased @xcite .
the situation concerning the predictions of inhomogeneous models @xcite is summarized in @xcite .
it is concluded that the inhomogeneities that would be able to amount to a @xmath31 effect as observed from supernovae must be subhorizon and of strong type and they require a treatment beyond first
order perturbation theory .
research in this field will continue towards quantitative predictions to improve the connection with observational cosmology @xcite .
the aim of discriminating between various possibilities for dark energy places stringent requirements on the use of standard candles @xcite . within the two decades from the discovery ( 1998 ) and the gathering of space mission data , it is feasible to reduce largely the errors , as mentioned before , and this would reflect into an error of only 1@xmath140 2@xmath140 in @xmath207 .
errors much bigger than this 1@xmath140 of error in d@xmath0 result in the impossibility to distinguish between models . even at the level of 1@xmath140 in d@xmath208
one finds some degeneracy @xcite .
space projects aimed at determining the nature of dark energy are counting on a systematic error in the method around 0.02 mag ( i.e. 1% in the distance ) @xcite .
a different question is whether from ground , within the planned experiments at work in the present decade , one can converge towards that limit .
these questions have been investigated in detail by using monte carlo simulations including the systematic errors in @xmath27 ( or @xmath207 ) to be found in the evaluation of the present value of the equation of state of dark energy @xmath126 and its first derivative in the scale factor @xmath170 @xcite ( see expression 33 ) .
snap is a mission that has been specifically designed to gather about 2000 sne ia up to z @xmath42 1.7 . as a single mission gathering supernova data with the same instrumentation
, it can eliminate possible photometric offsets and include high precision photometry out to 1.7 microns@xcite . with such large number of sne ia , statistical errors in every redshift bin of @xmath209 @xmath10 0.1 become negligible .
the error is dominated by the systematics . the total error in each redshift bin
is given by : @xmath210 with expected systematic error ; @xmath211 table i gives a progress evaluation in the control of high redshift supernova distance uncertainties .
table i + error contributions to high supernova distances in 1998 + [ cols="<,^,<,^",options="header " , ] -3pt 1 true cm @xmath212 0.00 * means values of the order of 10@xmath213 .
the statistical errors are for the redshift bin .
1 true cm @xmath214 in the table for prospects in a space mission , the statistical error from light curve sampling 1 true cm of sne ia is reduced due to the large number of sne ia data per redshift bin .
ground based projects will not be able to achieve such accuracy .
they aim at a reduction of systematic errors to 0.040.05 mag .
much of what will be feasible from now until the next space era of determination of the equation of state , depends on the use of all the information already gathered in the best way .
spectra gathered so far can be exploited further than what it has been done up to now , in order to reduce systematic errors . in the coming years
, a lot can be done to improve the present dispersion of the hubble diagram of sne ia and aim at a scatter of 1@xmath140 in d@xmath0 .
the above limit is only reachable if all the systematic effects are dealt with to the best .
the sne ia test of @xmath25 benefits from using information on @xmath61 obtained by some other method .
this can come from weak lensing or from large scale structure @xcite .
baryon acoustic oscillations can also provide a measurement of distance along z , in this case , angular distances @xmath215 and of @xmath216 @xcite . generally , for any geometry
, one has : @xmath217^{{2\over3}}\ ] ] a measurement of @xmath218 has been obtained @xcite .
@xmath219^{2/3}\ ; = \ ; 0.469\pm0.017\ ] ] where @xmath220 is the redshift at which the acoustic scale has been measured in the redshift sample . combining bao and sneia distances
one can aim at tracing dark energy all through a high z range .
the value of @xmath221 is provided by wmap and planck . in between , other methods such as studying the growth of large scale structure ( lss ) obtained from ly@xmath49 clouds and clusters would depict the growth factor of structure along z. considering the information at high z derived from lss , sne ia and cmb should lead to learn about early dark energy @xcite .
type ia supernovae can test gr against other alternative theories of gravitation .
theories of higher dimensional gravity have different friedmann equations from those originating from gr .
some of those modified gravity proposals have already been tested and ruled out through the hubble diagram of sneia up to very high z. if dark energy is due to a scalar field either minimally coupled to gravity ( i.e. quintessence ) or non minimally coupled to gravity ( arising from scalar tensor gravity or the effective low energy superstring proposals ) , the hubble diagram up to z @xmath174 1 obtained by the supernova collaborations gives the opportunity for exhaustive tests . in what concerns
scalar tensor gravity , supernovae are a precision experiment in a way analogous to the observations of the psr b1913@xmath3316 or to solar system tests .
moreover , supernovae have already provided hints on the form of the potential of the hypothetical scalar field ( for the case of minimally coupled field ) that could be associated to the acceleration of the universe . nonetheless , gr plus a form of cosmological constant is not ruled out observationally .
finding out how is @xmath134 close to @xmath222 is one of the most intriguing questions .
hopefully , this will be clarified by the samples to come .
the goal of obtaining a consistent picture on @xmath25 up tp z @xmath42 1.5 from the present samples is approached through the various sn procedures used by the existing collaborations .
ultimately , supernovae together with complementary methods just recently discovered shall select the correct ideas on dark energy and gravity .
if any modified gravity theory has a cosmological effect , those methods should start to see it .
riess a g et al .
( the high z sn search collaboration ) 1998 _ astron . j. _ * 116 * 1009 perlmutter s et al.(the supernova cosmology project ) 1999 _ apj _ * 517 * 565 einstein a 1917 _ sitzungsber .
142152 brans c h and dicke r h 1961 _ phys . rev _ * 124 * 925 will c 1993 _ theory and experiment in gravitational physics _ , cambridge ( london ) green m b , schwarz h and witten e. 1988 _ superstring theory _ ( cambridge university press , cambridge )
arkani - hamed n , dimopoulos n and dvali g 1998 _ phys .
_ b * 429 * 263 ( _ preprint _ hep
ph/9807344 ) dvali
g gabadadze g and porrati m 2000 _ phys lett b _ * 485 * 208 deffayet c dvali g and gavadadze g 2002 _ phys .
_ d * 65 * 044023 adelberger e g , heckel b and nelson a e 2003 _ ann .
nucl . part .
* 53 * 77 adelberger e g , heckel b and hoyle c d 2005 _ physics world _ * 18 * 41 adelberger e g et al .
2007 _ phys .
* 98 * 131104 ( _ preprint _ hep - th/0611223 ) kneller j p and steigman g 2003 _ phys rev _
d * 67 * 063501 koyama k and maartens r 2006 _ jcap _ * 601 * 016 maartens r 2004 _ living rev . rel . _
* 7 * 7 spergel d n et al .
( wmap collaboration ) 2003 _ apj supp _ * 148 * 175 ( _ preprint _ astro
ph/0302209 ) sahni v and starobinsky a 2006 int .
j. m. phys d * 15 * 2105 ( _ preprint _ astro
ph/0610026 ) calcagni g , de carlos b and de felice a 2006 _ nucl .
_ b * 752 * 404 ( _ preprint _ hep - th/0604201 ) de felice a hindmarsh m and trodden m 2006 _ jcap _ * 0608 * 005 ( _ preprint _ astro - ph/0604154 ) gorbunov d , koyama k and sibiryakov s 2006 _ phys .
_ d * 73 * 044016 ( _ preprint _ hep
th/0512097 ) padilla a 2007 ( _ preprint _ hep - th/0610093 ) trodden m 2006 in _ from quantum to cosmos : fundamental physics research in space , washington _
( _ preprint _ astro - ph/0607510 ) lukas a , ovrut b.a .
, stelle k.s and waldram d 1999 _ phys .
_ d * 59 * 086001 ( _ preprint _ hep - th/9803235 ) binetruy p , deffayet c and langlois d 2000 _ nucl .
b _ * 565 * 269 ( _ preprint _ hep - th/9905012 ) fairbairn m and goobar a 2005 _ phys .
b _ * 642 * 432 ( _ preprint _ astro - ph/0511029 ) maartens r and majerotto e 2006 _ phys .
_ d * 74 * 023004 ( _ preprint _
astro - ph/0603353 ) eisenstein d j et al .
2005 _ apj _ * 633 * 560 wang y and mukherjee p 2006 _ astrophys .
j _ * 650 * 1 ( _ preprint _ astro - ph/0604051 ) riess a g et al .
( the higher z team ) 2004 _ apj _ *
607 * 665 astier p et al .
( the snls collaboration ) 2006 _ a & a _ * 447 * 31 ( _ preprint _ astro - ph/0510447 ) dvali g and turner m 2003 _ fermilab - pub-03 - 040-a _ ( _ preprint _
astro - ph/0301510 ) randall l and sundrum r 1999 _ phys .
* 83 * 3370 randall l and sundrum r 1999 _ phys .
* 126 * 207 binetruy p , deffayet c , ellwanger u and langlois d 2000 _ phys .
_ b * 477 * 285 langlois d 2002 _ int j. mod phy _ a * 17 * 2701 ( _ preprint _
gr - qc/0205004 ) brax p and van de bruck c 2003 _ class .
grav . _ * 20 * 201 ( _ preprint _ hep
th/0303095 ) easson d a 2000 _ int .
_ a * 16 * 4803 durrer r 2005 , hep - th/0507066 weller j and albrecht a 2001 _ phys .
* 86 * 1939 ratra b and peebles p.j.e .
1988 _ phys .
d * 37 * 3406 wetterich c 1988 _ nucl .
b _ * 302 * 668 steinhardt p j , wang l and zlatev i 1999 _ phys . rev _ d * 59 * 123504 peebles p j e and ratra b 2003 _ rev . mod . phys _ * 75 * 559 copeland e j , sami m and tsujikawa s 2006 _ int .
_ d * 15 * 1753 ( _ preprint _ hep / th/063057 ) uzan j .-
p 2003 _ rev .
phys _ * 75 * 403 bertolami o and martins pj 2000 _ phys .
_ d * 61 * 064007 li c holz d and cooray a 2006 , astro ph/0611093 huterer t and turner m
s 1999 _ phys .
_ d * 60 * 1301 huterer t and turner m s 2001 _ phys .
_ d * 64 * 3527 sahln m liddle a r and parkinson d 2007 phys .
rev d * 75 * 023502 ( _
preprint _ astro - ph/0610812 ) knop r a , et al .
( the supernova cosmology project ) 2003 _ ap j _ * 598 * 102 riess a g et al .
( the higher z team ) 2007 _ apj _ * 659 * 98 brans c h 2005 in _ contributions to santa clara 2004 1st . international workshop on gravitation and cosmology _ ( _ preprint _ gr - qc/0506063 ) esposito - farese g. 2004 , in _ phi in the sky : the quest for cosmological scalar fields , aip conference proceeding _ * 736 * 35 amendola l 1999 _ phys . rev .
_ d * 62 * 043511 ( _ preprint _
astro - ph/9908023 ) amendola l and quercellini c 2003 _ phys .
_ d * 68 * 023514 ( _ preprint _ astro - ph/0303228 ) boisseau b. esposito farese g polarski d and starobinsky a 2000 _ phys
lett _ * 85 * 2236 ( _ preprint _
gr - qc/0001066 ) esposito - farese g and polarski d. 2001 , _ phys .
_ d * 63 * , 063504 perivolaropoulos l 2005 _ jcap _ * 0510 * 001 teyssandier p and tourrenc p 1983 _ j.math .
* 24 * 2793 gasperini m and veneziano g 2003 _ phys .
rept _ * 373 * 1 barrow j d and cotsakis s 1988 _ phys
_ b * 214 * 515 barrow j d and maeda k 1990 _ nucl .
b * 341 * 290 holden d j and wands d 2000 _ phys .
_ d * 61 * 043506 wands d 1994 _ class quant grav _ * 11 * 269 starobinsky a a 1980 _ phy lett _
b * 91 * 99 chiba t 2003 _ phys .
_ b * 575 * ( _ preprint _ astro - ph/0307338 ) capozziello s , carloni s. and trosi a. 2003 _ int . j. mod .
_ d * 12 * 1969 ( _ preprint _
astro - ph/0303041 ) carroll s m , de felice a and trodden m 2005 _ phys .
_ d * 71 * 023525 mena o , santiago j and weller j 2006 _ phys .
96 * sotirou t p 2006 _ class . quantum grav . _ * 23 * 5117 amendola l , polarski d and tsujikawa s 2007 _ phys .
lett _ * 98 * 131302 ( preprint astro - ph/0603703 ) ( astro - ph/0603703 ) antoniadis i 1992 in _ proceedings of string theory and quantum gravity , trieste 1992 _ 286 ( _ preprint _ hep - th/9211055 ) antoniadis i , rizos j and tamvakis k 1994 _ nucl .
_ b * 415 * 497 , ( _ preprint _ hep - th/9305025 ) neupane
i p 2006 _ class .
* 23 * 7493 gross d j and sloan j h 1987 _ nucl .
b * 291 * , 41 caldwell rr 2002 _ phys . lett .
_ b * 545 * 23 nojiri s , odintsov s d and sasaki m 2005 _ phys .
_ d * 71 * 123509 lue a and starkman g d 2004 _ phys .
_ d * 70 * 101501 chimento l p lazkoz r maartens r and quirs t 2006 _ jcap _ * 0609 * 004 bouhmadi
lopez m , gonzalez
diaz p f and martin
moruno p 2006 ( _ preprint gr - qc/0612135 _ ) koivisto t and mota d f 2007 _ phys .
b _ * 644 * 104 ( _ preprint _ astro - ph/0606078 ) ellis j , kalara s , olive k. a. and wetterich c 1989 _ phys .
_ b * 228 * 264 wetterich c 1994 _ aa _ * 301 * 321 ( _ preprint _ hep - th/9408025 ) brax p and martin j 2006 _ jcap _ * 0611 * 008 damour t piazza f and veneziano g 2002 _ phys . rev _
d * 66 * 046007 brax p and martin j 2006 ( _ preprint _ hep
th/0612208 ) hoyle cd et al .
2004 _ phys .
_ d * 70 * , 042004 murphy t w http://physics.ucsd.edu/ tmurphy / apollo / doc / multiplex.pdf doran m robberts g and wetterich c 2007 _ phys .
rev d _ * 75 * 023003 ( _ preprint _ astro - ph/0609814 ) witten e. 2000 _ lecture at dm2000 , marina del rey _
( _ preprint _
hep - ph/0002297 ) susskind l 2003 ( _ preprint _ hep
th/0302219 ) chevalier m and polarski d 2001 _ int j. mod .
_ d * 10 * 213 ( _ preprint _
gr - qc/0009008 ) linder e v 2003 _ phys .
lett _ * 90 * 091301 kim a g , linder e v , miquel r and mostek n 2004 _ mnras _ * 347 * 909 kowal c t 1968 _ aj _ * 73 * 1021 filippenko a v and riess a g 1998 _ phys
_ * 307 * 31 pskovskii y p. 1977
_ soviet astron .
_ * 21 * , 675 phillips m m 1993 _ apj _ * 413 * l105 phillips m m et al .
1999 _ aj _ * 118 * 1766 sullivan et al .
( the supernova cosmology project ) 2003 _ mnras _ * 340 * 1057 nugent p , kim a and perlmutter s 2002 _ pasp _ * 114 * 803 leibundgut b 2002 _ araa _ * 39 * 67 matheson t. et al .
2005 ( essence ) _ aj _ * 129 * 2352 hook i m et al ( the supernova cosmology project ) 2005 _ aj _ 130 2788 lidman c et al .
( the supernova cosmology project ) 2005 _ a a _ * 430 * 843 blondin s et al .
( essence ) 2006 _ aj _ * 131 * 1648 http://www.am.ub.es/ bcnde / darkeno.html riess a g , press w h and kirshner r p 1995 _ apj _ * 438 * l17 jha s riess a g and kirshner r p 2007 _ apj _ _ astrophys .
j _ * 659 * 122 perlmutter s et al .
( the supernova cosmology project ) 1998 _ nature _ * 391 * 51 goldhaber g et al .
( the supernova cosmology project ) 2001 _ apj _ * 558 * 359 tonry j.l .
et al . ( the high z sn search collaboration ) 2003 _ apj _ * 594 * 1 barris b j et al . ( the high z sn search collaboration ) 2004 _ apj _ * 602 * 571 clocchiatti a et al .
( the high z sn search collaboration ) 2006 _ apj _ * 642 * 1 freedman w l et al .
2001 _ apj _ * 553 * 47 sandage a et al .
2006 _ astrophys .
j _ * 653 * 843 ( _ preprint _ astro
ph/0603647 ) ruiz lapuente p 1997 _ astr .
ser . _ * 126 * 207 ( _ preprint _ astro
ph/9710019 ) ruiz lapuente p 1996 _ ap .
j. lett _ * 465 * 83 ( _ preprint _ astro
ph/9604044 ) goobar a and perlmutter s 1995 _ apj _ * 450*,14 garnavich
p m et al .
1998 _ apj _ * 509 * 74 wood vasey w m et al .
( essence ) 2007 ( _ preprint _ astro - ph/0701041 ) branch d , romanishin w and baron e 1995 _ apj _ * 465 * 73 ( _ preprint _ astro - ph/9510071 ) schlegel d. j. , finkbeiner m and davis m. 1998 _ apj.s _ * 500 * 525 perlmutter s 2005 _ hst proposal - go 10496 _ aldering g , knop r and nugent p 2000 _ aj _ * 119 * 210 conley a. et al . ( the snls collaboration ) 2006 _ aj _ * 132 * 1707 ivanov v d , hamuy m , and pinto p a 2000 _ apj _ * 542 * 588 quimby r et al .
( the supernova cosmology project ) 2002 _ baas _ * 201 * 2305 kim a g , goobar a and perlmutter s. 1996 _ pasp _ * 108 * , 190 amanullah r , mrtsell e and goobar a 2003 _ a & a _ * 397 * 819 jonsson j , dahlen t , goobar a , gunnarson c , mortsell e and lee k 2006 _ apj _ * 639 * 991 nesseris s and perivolaropoulos l 2007 ( _ preprint _ astro - ph/0612653 ) _ j. cosmol
. astropart .
phys . _ * 0702 * 025 ruiz
lapuente p et al .
2006 http://www.am.ub.es/ bcnde / darkeno.html ruiz lapuente p 2007 invited review at _
bernard s cosmic stories _ , valencia june 2006 rubin d et al .
( the supernova cosmology project ) 2006 _ baas _ * 209 * 9003r kowalski et al .
( the supernova cosmology project ) 2007 , in preparation buchert t , larena j and alimi j m 2006 _ class quant .
* 6379 bonvin c , durrer r and kunz m 2006 _ phys . rev . lett _ * 96 * 191302 discussions on inhomogeneous cosmologies at the conference _ key approches to dark energy _
barcelona 2006 .
this author thanks all the participants for addressing this topic .
rasanen s 2006 _ jcap _ * 0611 * 003 kolb e w , matarrese s , notari a and riotto a 2005 ( _ preprint _ astro - ph/0506534 ) clrier m
n 2000 _ astron
. astrophys _ * 353 * 63 buchert t 2007 gr - qc/0612166 maor i and brustein r 2003 _ phys .
d * 67 * 103508 espaa
bonet c and ruiz lapuente p 2005 ( _ preprint _ hep - ph/0503210 ) http://snap.lbl.org heavens a f , kitching t d and taylor a n 2006 _ mnras _ * 373 * 105 tegmark m et al .
2006 _ phys .
d _ * 74 * 123507 ( _ preprint _ astro
ph/0608632 ) percival w j et al .
2007 _ astrophys .
j _ * 657 * 645 ( _ preprint _ astro - ph/0608635 ) nichol r 2006 in _ cosmology , galaxy formation and astroparticle physics on the pathway to ska ( eds klckner h - r et al . ) _ yamamoto k bassett b a nichol r c and suto y 2006 _ phys rev d _ * 74 * 063525 ( _ preprint _ astro - ph/0605278 ) frieman j a , huterer d , linder e v and turner m s 2003 _ phys . rev . _
d * 67 * 083505 ( _ preprint _ astro - ph/0208100 ) wang y and freese k 2006 _ phys .
b _ * 632 * 449 dunkley j , bucher m , ferreira p g , moodley k and skordis c 2005 _ phys . rev .
lett . _ * 95 * 261303 kunz m , trotta r and parkinson d 2006 _ phys .
_ d * 74 * 023503 | the discovery of the acceleration of the rate of expansion of the universe fosters new explorations of the behavior of gravitation theories in the cosmological context .
either the gr framework is valid but a cosmic component with a negative equation of state is dominating the energy matter contents or the universe is better described at large by a theory that departs from gr . in this review we address theoretical alternatives that have been explored through supernovae .
* contents * \1 .
introduction \2 . gravity and extra dimensions \3 . the dark energy field and scalar tensor gravity \4 .
the cosmological constant ?
\5 .
the d@xmath0 test from high z sne ia \6 .
testing the adequacy of the frw metric \7 .
next decade experiments \8 .
complementarity \9 . |
phase anisotropy together with fluidity are the basic unifying features of all liquid crystals .
fluidity implies some disorder of the molecular positions in one or more spatial dimensions .
phase anisotropy reflects the directionality of the molecular organisation in these systems and can be of the a - polar type , if each direction in the bulk phase is physically indistinguishable from its opposite , or of the polar type , if such indistinguishability does not hold for all directions . in the latter case the underlying molecular organisation will exhibit polar orientational ordering .
accordingly , the bulk phase could in this case acquire spontaneous electric or magnetic polarisation if the molecules have permanent electric or magnetic dipoles . aside from their purely scientific interest ,
the many actual and potential technological applications of liquid ferroelectric and ferromagnetic systems have stimulated intensive research on polar liquid crystals [ 1 ] . in this paper
we discuss various molecular mechanisms that have been proposed for the understanding of polar ordering in the most common liquid crystalline phases .
we consider low molar mass nematic , smectic and columnar liquid crystal phases and we concentrate on ( i ) the molecular interactions that could promote or oppose polar organisation and ( ii ) the quantification of polar orientational ordering in terms of order parameters and of macroscopic physical properties .
our considerations are restricted to electric properties and interactions but all the theoretical arguments can be carried over to magnetic systems as well .
a polar nematic would be a translationally uniform phase with long - range polar orientational order .
the most symmetric such phase is a uniaxial polar nematic , i.e. a nematic with full rotational symmetry about a unique phase - fixed axis denoted by the unit vector @xmath0 , the director , but with broken `` up - down '' symmetry so that the directions @xmath1 and @xmath2 are not equivalent .
the minimal molecular asymmetry requirements for the formation of such a phase correspond to uniaxially polar molecules , i.e. polar molecules with an axis @xmath3 of full rotational symmetry , with the polarity rendering the molecule asymmetric with respect to the direction reversal @xmath4 . due to this asymmetry
the ensemble average @xmath5 of the molecular unit vector @xmath3 does not vanish .
the single - molecule probability distribution is in this case a function of just one molecular variable , namely @xmath6 , where @xmath7 is the angle of the molecular unit vector @xmath8 relative to the director @xmath9 .
the moments of this distribution are conveniently represented by the ensemble averages ( order parameters ) @xmath10 of the legendre polynomials of rank @xmath11 .
the quantification of polar ordering for such a system is given by the ensemble average @xmath12 .
the primary measure of the magnitude of polarity is thus the first rank order parameter @xmath13 .
the second rank order parameter @xmath14 , which is the leading order parameter for common ( a - polar ) nematics , is related to the breadth of the polar distribution according to the relation @xmath15 from which it becomes apparent that @xmath13 places the following lower bound on @xmath14 , @xmath16 if the molecules forming the uniaxial polar nematic phase carried permanent electric dipole moments @xmath17 with a nonvanishing component @xmath18 along the direction of the molecular symmetry axis @xmath8 , then the phase would exhibit spontaneous electric polarisation @xmath19 where @xmath20 denotes the molecular number density and @xmath21 .
thus , the spontaneous polarisation vector would be in the direction of the phase symmetry axis @xmath22 and proportional to the polar order parameter @xmath23 according to the relation @xmath24 this equation relates , via a molecular property @xmath25 , a macroscopic electrostatic measure of polarity @xmath26 to a microscopic measure of polarity represented by the `` vector order parameter '' @xmath27 .
there does not seem to be any thermodynamic or symmetry argument forbidding _ a priory _ the formation of a polar nematic phase , uniaxial or not . yet , to date none of the existing low molar mass nematogens is known to exhibit such a phase[2 ] .
it thus becomes theoretically challenging to rationalise , on the molecular level , the a - polarity of common nematics and from there to ask what sort of molecular interactions would be required to stabilise polar ordering in the nematic phase . for example , permanent electric dipole moments of a few debyes are quite usual for molecules forming nematic phases and have measurable effects on their ( a - polar ) orientational order [ 3 ] .
an obvious question is then whether electric dipole interactions could bring about phase polarity in nematics [ 4 ] .
the answer is negative .
a simplified explanation can be sketched as follows .
the interaction potential between two dipoles , @xmath28 and @xmath29 is given by @xmath30 / r^3\ ] ] where @xmath31 and @xmath32 are the interdipole unit vector and distance respectively .
consider the condensed , positionally disordered , phase of dipolar spherical molecules ( i.e. molecules whose interactions in excess the dipole - dipole interaction are isotropic ) . if short - range position - orientation correlations are completely ignored then the contribution of the electrostatic interaction to the free energy for the phase with polar orientational order ( fig .
1a ) vanishes .
this is a reflection of the particular dependence of @xmath33 on @xmath31 for fixed dipole directions and interdipole distance .
the most directly relevant feature of this dependence is illustrated in fig .
2a , which shows that the energy of `` side - by - side '' configurations of parallel dipoles is half the magnitude and of opposite sign relative to the energy of the `` head - to - tail '' configurations .
since in three dimensions there are two side - by - side configurations for each head - to - tail , the average dipole - dipole energy , in the absence of correlations , vanishes . the same null result is obtained when averaging the dipole - dipole contribution to the free energy in the orientationally disordered phase ( fig 1b ) , which however has larger entropy and will therefore always be thermodynamically more stable than the polarly ordered state .
essential to these arguments is the neglect of short - range correlations .
the latter , however , are not always negligible .
in fact , simulations of hard dipolar spheres have produced , for sufficiently strong dipoles and low densities , stable phases of polar strings , ( linear aggregates of dipolar spheres in head to tail succession of the dipole moments , fig .
1c ) [ 5,6 ] .
the correlations in these strings are such that the head - to - tail configuration dominates at short distances whereas the side - by - side ( parallel or antiparallel ) configurations correspond to molecular pairs on different strings , which are therefore more distant .
this leads to an electrostatic reduction of the free energy that could compensate for the entropy reduction associated with the formation of the polar strings and thus provide thermodynamic stability relative to the completely disordered phase . on increasing the density ,
the side - by - side configurations are forced to the same average proximity as the head - to - tail and the chains eventually decompose into a positionally disorder a - polar fluid .
if instead of dipolar spherical molecules one considers elongated molecules with longitudinal dipoles then the polarly ordered condensed phase is further destabilised with respect to the a - polar phase .
this happens because now , due to larger separations of the dipoles , the magnitude of the head - to - tail energy is less than twice the magnitude of the side by side energy ( fig .
2b ) and therefore the positionally uncorrelated averaging will yield a net increase of the free energy of the polar phase relative to the a - polar one .
thus the stronger the dipole the more it disfavours the polar nematic ordering of the elongated molecules and in fact turns out to favour the a - polar smectic a phase[7 ] . in the case of disc - like molecules with central dipoles along the disc symmetry axis , the electrostatic contribution favours the polar ordering .
this is so because the head - to - tail energy , due to smaller interdipole separations , is now more than twice the magnitude of the side - by - side energy ( see fig . 2c ) .
as shown in monte carlo ( mc ) simulations , however , the organisation of the polar discs does not correspond to a polar nematic but rather to a columnar phase with polar ordering within individual columns and antiferroelectric configurations of the columns [ 8 ] .
it becomes apparent from these considerations that the formation of a polar nematic phase is not favoured by strong longitudinal electric dipoles .
it rather requires interactions where the polar disposition of the molecules can be favoured but without forcing them to into any kind of positional register that would destroy the uniform distribution of their positions .
interactions of the amphiphilic type , for example , can favour the polar alignment of elongated molecules , as will be discussed below , but only as a result of ( in the present context , at the expense of ) phase microsegregation [ 9 ] .
furthermore , steric repulsions originating from the polar shape - asymmetry of tapered molecules , wedge shaped , pear shaped , etc , disfavours polar ordering in the undisturbed bulk phase since such shapes pack more efficiently in antiparallel configurations .
it thus seems that none of the basic interactions encountered in common liquid crystals , nor any simple combination thereof , favours nematic polar ordering .
this explains , at least partly , why a low molar mass nematic phase has not been detected experimentally to date .
recently , berardi , ricci and zannoni [ 10 ] where successful in tailoring a model potential that can produce , under certain parameterisation , a polar nematic phase in mc simulations .
their model interaction essentially behaves in the opposite way to the dipole - dipole interaction ( fig .
2b ) in that it assigns the lowest energy to the side - by - side parallel arrangement of a molecular pair , high energy to the parallel head - to - tail and intermediate energy to the side - by - side antiparallel . of course the identification of real molecules that would interact in the required way is the major step to be taken . in the present state of molecular design efforts ,
however , it appears that much more is known on what should be avoided than on what is to be pursued .
polar ordering in orthogonal smectics is quite usual [ 11 ] . known variants of the smectic a phase ( _ sm - a _ ) with polar layers include the bilayer phases @xmath34 , @xmath35 and _ @xmath36 _ ( fig .
3a , b , c ) . in these phases
the polar asymmetry is exhibited in the direction of the layer normal , with adjacent layers having opposite polarities thus rendering the bulk phase macroscopically a - polar .
on the molecular scale , the structure of the phases @xmath35 and _ @xmath37 _ suggests that the molecular interactions giving rise to the polar ordering should be such as to favour the side - by - side parallel alignment of molecules within the same layer while favouring the head - to - head or tail - to - tail alignment of molecules in consecutive layers ( fig .
obviously neither the electric dipole forces nor the purely steric forces associated with wedge - type shape asymmetry favour such configurations ; in fact they strongly disfavour them .
in contrast , amphipile type forces favour both types of configurations and are therefore considered the primary interaction underlying the formation of these phases .
the bilayer structure of the @xmath34 phase suggests molecular interdigitation among adjacent sublayers such that the favoured molecular arrangements produce side - by - side antiparallel configurations of the molecular tips ( fig .
these configurations could be favoured by interactions from dipole moments situated near the ends of the elongated molecules and in fact theory and simulations show that model rod - like molecules with off centre axial dipole moments can produce the @xmath34 phase structure [ 12 ] .
however , the amphiphilic interactions also favour this type of configurations .
moreover , this arrangement makes the packing more effective in the presence of wedge like shape asymmetry , particularly in combination with amphipilicity .
this is then an example of ordering where several types of interactions could separately or cooperatively produce the polarity .
in fact , photinos and saupe [ 13 ] showed nearly three decades ago the theoretical possibility of longitudinally polar smectics based on a generic molecular field description of the polar interactions .
the type of polarity described above for orthogonal smectics is similar to that of nematics in that it consists of the local ( within a sublayer ) breaking of the up - down symmetry @xmath38 along the director and entails the same minimal asymmetry of the molecular structure , i.e. uniaxial molecules with a polar axis @xmath39 of full rotational symmetry .
consequently , the polarity in each sublayer is quantified by the vector order parameter @xmath40 as in uniaxial polar nematics . in the presence of longitudinal dipoles a spontaneous electric polarisation will be produced in each sublayer according to eq .
( [ eq3 ] ) and , due to the alternating polarity of the bilayer structure , the resulting `` antiferroelectric '' system will appear electrically a - polar on the bilayer scale .
the term longitudinal will be reserved for the above type of polarity , namely the one that manifests itself as the breaking of the @xmath38 symmetry along the director ( coincident with the layer normal in the case of orthogonal smectics ) , in order to distinguish it from the transverse polarity that is manifested as the breaking of the twofold rotational symmetry about the director * n*. ( b ) semibilayer _ _ sm - a__@xmath41 .
( c ) ribbon bilayer _ sm-@xmath36_. ( d ) banana - type _ _ sm - a__@xmath42 , shown here in antiferroelectric succession of the layers .
( e ) rod - like polar smectic a phase model , shown here in antiferroelectric succession of the layers .
( f ) banana - type orthogonal smectic phase model with both , longitudinal ( along the layer normal ) and transverse polarity , shown here in antiferroelectric succession of the layers with respect to both polarities .
( g ) rod - like orthogonal smectic phase model with both , longitudinal ( along the layer normal ) and transverse polarity , shown here in antiferroelectric succession of the layers with respect to both polarities .
the arrows attached to the molecules in figures ( d ) to ( g ) indicate the molecular vectors @xmath43.,width=288 ] an orthogonal smectic phase with transverse polarity and positional disorder has been reported recently [ 14 ] for compounds with banana - type molecular shape asymmetry .
the polarity of the layers in this case is thought to result from the stacking of the molecules so that the `` steric dipoles '' associated with their bent shape become aligned ( see fig .
this polar stacking can produce a spontaneous polarisation in the plane of the layers if the molecules possess electric dipole moments with nonvanishing component along the direction of the steric dipole .
it should be noted that such electric dipole moments , if present , would contribute to the polar alignment because their configuration within each layer gives on average equal numbers of parallel side - by - side and head - to - tail pairs while the side - by - side pairs belonging to different layers are more distant and thus an overall reduction of the internal electrostatic energy relative to the a - polar system would result .
in fact , unlike the case of longitudinal polarity , transverse polarity can in principle be generated by electric dipole interactions alone .
consider for example idealised smectic layers consisting of a planar arrangement of positionally disordered parallel rod like molecules with central transverse electric dipoles ( fig .
3e ) . the internal electrostatic energy per molecule for the state with perfect alignment of the dipoles is lower than that of the state with randomly oriented dipoles by @xmath44 , where @xmath45 is the dipole strength , @xmath46 is the surface density of the molecules in the layers and @xmath47 is the diameter of the rods ( the distance of closest approach of the dipoles , fig 2d ) .
the entropy per molecule for the polar state is lower than that of the random state by @xmath48 and therefore the polar state becomes the thermodynamically stable one for sufficiently large values of the quantity @xmath49 . at this level of approximation
the intra - layer electrostatic interactions do not contribute to the internal energy of the system since the interaction energy between any dipole and all the dipoles belonging to a different layer averages to zero , both in the polar and the random states .
this result suggests that , to a first approximation ( i.e. ignoring orientational and positional fluctuations and correlations of inter- and intra - layer molecular motions ) , the inter - layer electrostatic interactions are indifferent to the mode of propagation of polarity across the layers ( parallel or antiparallel disposition of adjacent layers ) .
monte carlo simulations of rod like molecules with transverse dipoles give some indications of in - plane polar order [ 15 ] .
gil - villegas , mcgrother and jackson [ 16 ] found that , even at fairly high densities , the dipoles in a layer self - organise to form ring domains that evolve , on lowering the temperature , into antiferroelectric elongated chain domains .
it is apparent from the examples of banana shaped molecules and of rods with transverse dipoles in fig .
3d , e that the minimal molecular asymmetry required for in - plane phase polarity corresponds to molecules with a plane of symmetry and a twofold symmetry axis @xmath43 on that plane .
the molecular `` long axis '' @xmath8 is defined as the axis within the plane of symmetry and perpendicular to @xmath43 and is no longer an axis of rotational symmetry .
the molecular polarity consists in the lack of @xmath50 symmetry .
accordingly , phase polarity can be quantified by the vector order parameter @xmath51 .
the direction of @xmath51 defines a second director @xmath52 in the plane of the layer and thus perpendicular to the primary director @xmath22 .
consequently , transverse polarity makes the orthogonal smectic phase necessarily biaxial .
the asymmetry associated with in - plane polarity consists in the breaking of the @xmath53 symmetry while maintaining the @xmath54 symmetry .
when both symmetries are valid one has a biaxial ( non - polar ) orthogonal phase whereas the simultaneous breaking of both symmetries yields an orthogonal smectic with both in - plane and longitudinal polarity ( see fig .
the types of molecular interactions that can produce the combined polarity can in principle be furnished by the superposition of the interactions giving rise to the individual polarities , for example amphiphilic interactions for the longitudinal polarity and banana type steric dipole for the transverse polarity ( fig .
3f ) , since such interactions are not incompatible or counteracting .
it thus can be seen that the consideration of the two types of polar ordering and of their combination broadens the variety of orthogonal smectics with positionally disordered layers .
these possibilities of polarity are carried over and further enriched for orthogonal smectics with in - plane
positional order , the discussion of which is beyond the present scope .
the least ordered of the tilted smectic phases is the _ sm - c _ phase .
it consists of positionally disordered layers in which the orientational order of the molecules defines a director @xmath22 forming an angle ( the tilt angle ) with the direction normal to the layers .
the _ sm - c _ layers have a plane of symmetry ( the tilt plane , formed by the layer normal * z * and the director * n * ) .
they also have a twofold symmetry axis @xmath55 perpendicular to the plane of symmetry and a centre of inversion at the intersection of the twofold axis with the symmetry plane .
the directions @xmath22 and @xmath56 are equivalent ( physically indistinguishable ) . however , the structure of the _ sm - c _ phase is inherently polar since the tilt singles out a unique direction about the layer normal .
this unique direction is often represented by the so - called * c*-director [ 17 ] whose direction is defined from the projection of the * n*-director onto the layer plane but with @xmath57 and @xmath58 describing physically distinct states . a particularly useful alternative [ 18 ] representation is provided by the tilt pseudovector * t * defined in terms of the layer normal * z * and the * n*-director according to the relation @xmath59 obviously * t * is perpendicular to the tilt plane , it is invariant with respect to the replacement of * n * by *n * and the states * t * and *t * describe layers of opposite tilt and are therefore physically distinct .
is compatible with mirror symmetry of the layer ; the two mirror images are identical .
( c ) tilted smectic layer with chiral molecules carrying transverse electric dipole moments .
the resulting spontaneous electric polarization @xmath60 breaks mirror symmetry in the layer ; the two mirror images are distinct.,width=297 ] the polarity of the _ sm - c _ phase , being clearly a direct consequence of the tilted ordering , is referred to as indigenous polarity [ 19 ] .
it is of fundamentally different nature from the polarity that could appear in orthogonal smectics or in nematics .
first , when polar ordering appears in orthogonal smectics or in nematics , mirror symmetry perpendicular to the polar director ( see fig .
4a ) is broken .
this type of polar ordering is referred to as vector - type polarity , in accordance with the sign reversal of vectors under mirror reflection .
in contrast , the indigenous polarity of the _ sm - c _ phase is compatible with mirror symmetry in the plane normal to the direction of the polar asymmetry ( the tilt plane ) as illustrated in fig . 4b .
by analogy to the invariance of the direction of pseudovectors under mirror reflection , the indigenous polarity of the _ sm - c _ layers can then be termed as pseudovector - type . in order to quantify the indigenous pseudovector polarity of the _ sm - c _ layers and to illustrate its molecular origins we consider an idealised molecular model embodying the minimal asymmetry required for the appearance of tilted orientational ordering and positional disorder within the layers . for simplicity
we consider a rigid molecule whose shape has the same symmetries as the _ sm - c _ phase ( equivalently , the minimal asymmetry compatible with the _ sm - c _ phase ) .
a convenient example is the oblique cylinder of fig .
5a . it has a plane of symmetry , a twofold axis and an inversion centre .
objects of this shape have been used as molecular models in a theory of the _ sm - c _ phase by somoza and tarazona [ 20 ] . in the present context
they are used merely in order to convey the molecular symmetry by means of simple images .
required for the formation of the sm - c phase .
the vectors @xmath61 are on the oblique - cylinder plane of symmetry , to which the pseudovector @xmath62 is perpendicular .
the possible presence of a dipole moment @xmath28 directed out of the @xmath63 plane would render the structure chiral .
( b ) opposite orientations of a are equally probable in the sm - a phase .
( c ) the tilted molecular arrangement in the sm - c layer favours one of the two opposite orientations of a perpendicular to the tilt plane , thus generating psedovector indigenous polarity.,width=288,height=198 ] let the vectors * s * and @xmath64 be directed along the cylinder axis and normal to the oblique surfaces respectively .
the two halves on either side of the symmetry plane of the molecule ( the two `` faces '' of the molecule , for brevity ) are distinct mirror images of each other .
the pseudovector @xmath65 can then be used for the distinction between the two faces . in
the _ sm - a _ phase there is clearly nothing to make any given orientation of @xmath66 have a different probability from its opposite ( fig . 5 b ) whereas the tilted ordering of the _ sm - c _ phase induces a preference of one of the two opposite orientations perpendicular to the tilt plane over the other ( fig .
this preference gives rise to a nonvanishing value of the ensemble average @xmath67 and allows the definition of the ( pseudovector ) order parameter @xmath68 that quantifies the indigenous polarity .
clearly , the indigenous polarity is not related to chirality .
this is depicted in fig .
4b where it is apparent that the presence of pseudovector @xmath69 does not differentiate a layer from its mirror image .
however , if the indigenously polar _ sm - c _ phase is to exhibit spontaneous electric polarisation , it is necessity that the molecules possess an electric dipole moment with a nonvanishing component along @xmath70 . in that case
the indigenous polarity @xmath69 will generate a spontaneous electric polarisation @xmath26 normal to the tilt plane according to the relation @xmath71 where @xmath72 is a pseudoscalar measure of the strength of what is often referred to as the `` transverse dipole moment '' .
it is obvious , however , that the attachment of a transverse dipole component to the molecules ( see fig .
5a ) destroys the mirror plane symmetry of their structure , which thus becomes chiral . in this sense ,
the molecular parameter @xmath73 provides a measure of the electrostatic chirality of the molecular structure .
the electrically polar layer is no longer identical to its mirror image ( see fig .
4c ) , i.e. the layers have become electrostatically chiral .
according to the above considerations , the relevance of vector and pseudovector polarity to spontaneous polarisation can be summarised by stating that vector polarity is not necessarily present and , when present , can give rise to spontaneous polarisation even if the molecules are not chiral whereas pseudovector polarity is necessarily present in all tilted smectics but can give rise to electric spontaneous polarisation only if the molecules are chiral . this difference is born by equations ( [ eq3 ] ) and ( [ eq6 ] ) , which are formally identical but the two different mechanisms by which @xmath74 is generated are explicitly reflected on \(i ) the different physical significance of the molecular quantities @xmath75 ( scalar measure of longitudinal dipole ) and @xmath76 ( pseudoscalar measure of chirality - related transverse dipole ) and \(ii ) the molecular order parameters @xmath77 and @xmath69 quantifying respectively the direct ( vector ) polarity and the indigenous , tilt - generated , pseudovector polarity . having used an explicit molecular idealisation ( fig .
5 ) to illustrate the significance of the ( pseudovector ) indigenous polarity , its role in giving rise to spontaneous electric polarisation in tilted smectics , its differentiation from the latter polarisation and from the ( vector ) polarity appearing in orthogonal smectics and nematics , it is useful to give a generalised description of the above in terms of molecular and phase symmetry : the _ sm - c _ layers have @xmath78 symmetry .
the plane of symmetry is identified with the tilt plane and the twofold symmetry axis @xmath55 is perpendicular to that plane . as a result of the rotational symmetry about this axis ,
the projection of any vector or pseudovector order parameter onto the tilt plane necessarily vanishes .
furthermore , the plane of symmetry makes all vector order parameter components along the @xmath79 axis vanish but allows the appearance of pseudovector components along that axis . to exclude
the possibility of such pseudovector components would require at least one symmetry axis in the tilt plane , but such symmetry axis is of course precluded by the monoclinic symmetry of the tilted layers .
hence , tilt generates pseudovector indigenous polarity @xmath80 along the @xmath79axis of the _ sm - c _ layers . to identify the molecular origin of this polarity
it is necessary to consider molecular symmetry . to this end , it is sufficient to consider molecules that are not less symmetric ( in the statistical sense , when referring to rigid molecules ) than the phase itself , i.e molecules with @xmath78 symmetry . again , the existence of a plane of symmetry together with the lack of any symmetry axes on that plane makes it possible to define only molecular pseudovectors directed perpendicular to the plane of symmetry , i.e. along the molecular @xmath55 axis .
the ensemble average of the projection of such a pseudovector of unit length along the @xmath55 axis of the layers defines the order parameter @xmath81 .
the presence of the molecular @xmath55 axis is not essential to these considerations ; only the molecular plane of symmetry is necessary in order to exclude molecular chirality .
now , if chirality is allowed for in the smectic layers then there will be no symmetry plane .
the chiral analogue of the smectc - c phase , the _ sm - c * _ , has only one @xmath55 axis .
this precludes the appearance of any vector or pseudovector order parameter components in the tilt plane ( defined as the plane perpendicular to the @xmath55 axis , but no longer a symmetry plane ) .
it allows , however , the appearance of both vector and pseudovector components in the @xmath55 direction .
this leads to the possibility of simultaneous ( pseudovector ) indigenous polarity and ( vector ) spontaneous polarisation in chiral tilted smectics .
combinations of pseudovector polarity with the vector polarity described in the previous section for orthogonal smectics are possible and generate several polar variants of the _ sm - c _ phase .
thus , longitudinal vector polarity is present , in addition to indigenous pseudovector polarity , in the tilted analogues _ _ sm - c__@xmath82 , _ _ sm - c__@xmath41 and _ _ sm-__@xmath83 ( fig . 6 a , b ,
c ) of the orthogonal bilayer variants of the _ sm - a_. the tilted sublayers in these phases are polar along the director , i.e. have broken @xmath84 symmetry , and are therefore lacking the twofold symmetry axis ( normal to the tilt plane ) of the common _ sm - c _ layers
. however , this symmetry is restored in the bilayers due to the opposite longitudinal polarity @xmath85 of their constituent sublayers . as in the case of orthogonal smectics , these sublayers could exhibit spontaneous electric polarisation @xmath86 along the director ( consequently in the tilt plane and tilted with respect to the layer normal ) . for chiral molecules with transverse dipole moments this component of @xmath26
would be superposed to the component ( perpendicular to the tilt plane ) associated with the indigenous polarity and thus the total @xmath26 in each sublayer would be in the plane containing the director @xmath22 and the tilt pseudovector @xmath87 .
is directed normal to the to the tilt plane .
, width=326 ] transverse vector polarity can be combined with indigenous pseudovector polarity in many ways and results in the various banana - type tilted smectics .
thus , the vector order parameter @xmath88 can be in the tilt plane ( therefore perpendicular to the direction of the indigenous polarity @xmath89 , resulting in the layer structure of the @xmath90 phase ( fig .
6d ) , or it can be perpendicular to the tilt plane ( therefore parallel to @xmath91 , resulting in @xmath92 layers fig .
6e , or it could be in any intermediate direction ( within the plane perpendicular to the director and to the tilt plane ) resulting in @xmath93 layers [ 21 ] . finally , the possible polymorphism of polar tilted smectics could be further enriched by combining the indigenous pseudovector polarity @xmath69 , simultaneously with the transverse , @xmath51 , and the longitudinal , @xmath94 , vector - type polarities ( fig .
polar ordering , both of the vector and pseudovector type , is possible in columnar liquid crystals [ 22 ] . by analogy to smectic layers , the columns in orthogonal columnar phases can exhibit longitudinal vector polarity , with the order parameter @xmath77 directed along the column axis ( fig .
7a , b ) , or transverse vector polarity , with the vector order parameter @xmath95 directed perpendicular to the column axis ( fig .
for flat plate - like molecules , the longitudinal polarity can be generated by polar intermolecular interactions along the plate axis ( fig .
7a ) , including electric dipole interactions [ 8 ] .
the latter possibility obtains because , as shown in fig .
1c , the dipoles in head - to - tail configuration are separated by the `` thin '' dimension of the plate and thus the energy of that configuration becomes much lower than that of side - by - side . for bowl - shaped molecules ( fig .
7b ) , their polar ordering in each column is driven by their shape - dictated directional stacking .
transverse polarity can be generated in columns by amphiphilic interactions resulting from the partitioning of the plate into sectors of different philicity as shown in fig .
7c . in this case
the up - down symmetry of the column is maintained but the rotational symmetry about the column axis is lost and the thus columns , by becoming transversely polar , become inevitably biaxial .
columns exhibiting both longitudinal and transverse polarity have neither up - down symmetry nor rotational symmetry about the column axis and could be generated by combining the interactions corresponding to fig .
7a or 7b with those of fig .
7c , or , for example , by off - centre polar interactions directed across the plates , as shown in fig . 7d . in tilted columnar phases [ 23 ] the director @xmath96 , i.e. the direction of alignment of the molecular plate - normals , does not coincide with the direction @xmath97 of the axis of the column .
these two distinct directions define the tilt plane of the column .
the respective tilt pseudovector @xmath87 is defined in terms of @xmath22 and @xmath97 as in eq .
( [ eq6 ] ) .
the tilted ordering of the columns inflicts a pseudovector type polarity on the structure .
the molecular origin of the indigenous pseudovector polarity for tilted columnar liquid crystals can be illustrated in close analogy with that of tilted smectics [ 24 ] by considering idealised molecules of oblique disc shape as in fig . 8a .
the two molecular vectors @xmath98 and @xmath64 define a pseudovector @xmath99 whose direction differentiates between the two mirror image halves , or `` faces '' , of the oblique disc and whose ensemble average @xmath67 in the tilted column defines the indigenous polarity order parameter @xmath100 ( fig 8c ) .
the asymmetry is removed in an orthogonal column ( fig .
8b ) making @xmath101 .
the direction of @xmath102 is along the tilt pseudovector @xmath87 , i.e. perpendicular to the tilt plane of the column , and is therefore compatible with the two - fold rotational symmetry about @xmath87 .
moreover , the pseudovector character of @xmath69 makes it compatible with the mirror plane symmetry of the tilted columns . as in the case of tilted smectics , pseudovector polarity in tilted columnars
can be combined with longitudinal vector polarity ( along the director ) or with transverse vector polarity , or with both , to provide polar columns in which either the mirror plane symmetry or the twofold rotation symmetry or both symmetries are broken .
are on the oblique disc plane of symmetry , to which the pseudovector @xmath65 is perpendicular .
( b ) opposite orientations of * a * are equally probable in an orthogonal column .
( c ) the accommodation of the oblique disc in a tilted column favours one of the two opposite orientations of * a * perpendicular to the tilt plane , thus generating psedovector indigenous polarity.,width=297 ]
strictly , none of the known low molar mass liquid crystals exhibit , in their thermodynamically equilibrated bulk state , spontaneous macroscopic polarisation , electric or magnetic . the repetitive building blocks ( layers , columns , etc ) of phases with partial positional order are often polar but arranged in polarity neutralising patterns ( bilayer , antiferroelectric , helical , etc ) to produce macroscopically a - polar phases . in this paper
we have considered the different types of polar self - assembly of the building blocks , we have quantified the resulting polar ordering in terms of molecular order parameters and we have identified the underlying molecular symmetries and intermolecular interactions .
vector - type and pseudovector - type polarity refer to two distinct asymmetries of molecular ordering .
vector type applies when there is no plane of symmetry and no symmetry axis perpendicular to the direction of polarity .
pseudovector - type applies when there is no symmetry axis perpendicular to the direction of polarity but there is a plane of symmetry .
vector polarity is theoretically possible for nematics but requires interactions of rather unusual position - orientation dependence . in contrast
, amphiphilic , steric and electrostatic dipole interactions can , separately or cooperatively , give rise to vector polarity in the repetitive units of smectic and columnar phases .
the polarity in these cases could appear along the primary direction of alignment ( longitudinal ) or in a direction perpendicular to it ( transverse ) or both .
pseudovector polarity is produced by the tilted molecular arrangement in all a - chiral tilted smectic and columnar phases .
it is driven by the same interactions that give rise to the tilted ordering and reflects the asymmetry imposed on the molecular orientations by the packing constraints .
introducing polarity , with its different types and combinations , to the common a - polar modes of molecular self - organisation enriches the polymorphism of mesophases with many polar variants .
a good understanding of the molecular asymmetries and interactions that favour particular types of polar ordering is certainly essential to the design of such materials .
however , in addition to these qualitative assessments , quantitative estimates could be crucial to the successful design .
this is so because the thermodynamic stability of mesophases usually rests on a delicate balance between order - promoting interactions that are counteracted by entropy lowering .
therefore , the interactions need to be strong enough to produce the required molecular correlations but not so strong as to destabilise liquid crystallinity in favour of a solid phase . at present ,
theory and computer simulations could provide useful quantitative input but there are also many complications in obtaining realistic estimates .
for example , almost all liquid crystal forming molecules are flexible . for qualitative considerations molecular symmetry
is understood to refer to the `` conformationally averaged '' molecule and so are the intermolecular interactions . in a quantitative treatment , however , the individual molecular conformations ( normally lacking any element of symmetry ) have to be considered together with the detailed interactions of the various molecular segments .
this usually introduces an enormous number of degrees of freedom . assuming that the latter can be handled computationally , the predictive quality of the results rests on the detailed modelling of the intra- and inter - molecular interactions .
a reasonably realistic modelling of interactions in liquid crystals is in general very difficult due to the subtlety of these interactions .
for example , interactions associated with partial charge distributions are routinely modelled in terms of fixed electric dipole moments , quadrupole moments etc .
this kind of crude modelling does not allow for the possibly significant deformations of the charge distribution caused by intermolecular interactions ( polarisability ) .
it also disregards the fact that the multipole expansion of a charge distribution is valid only at distances that are large compared to the spatial extent of the distribution . as a result
the respective quantitative predictions would be at best indicative of the possible behaviour of the real system .
this paper is dedicated to the memory of pier luigi nordio . partial financial support for this work came from the eu _ tmr _ ( contract fmrx - ct97 - 0121 ) and through the funding of the polymer science and technology graduate studies programme of the university of patras .
recent reviews can be found in b. groh and s. dietrich , _ new approaches to problems in liquid state theory _ , ( kluwer , 1999 ) , pp .
173 - 196 ; s.t .
lagerwall , _
j. phys : condens . mat_.
* 8 * , 9143 ( 1996 ) ; p.i.c .
teixeira , j.m .
tavares and m.m.t . da gama , _
j. phys : condens
* 12 * , r411 , ( 2000 ) ; s.t .
lagerwall , _ ferroelectric and antiferroelectric liquid crystals _ ,
( john wiley & sons , ny , 1999 ) ; i. musevic , r. blinc and b. zeks , _ the physics of ferroelectric and antiferroelectric liquid crystals , _ ( world scientific , 2000 ) .
technically , liquid metals exhibiting ferromagnetism in the undercooled state ( t. albrecht , c. buhrer , m. fahnle , k. maier , d. platzek , and j. reske , _ appl . phys . _ a * 65 * , 215 ( 1997 ) ; a.n .
grigorenko p.i .
nikitin , a.y .
toporov , a.m. ghorbanzadeh , a. perrone , a. zocco and m.l .
de giorgi , _ appl .
phys . lett_. * 72 * , 3455(1998 ) ) could be termed as polar nematics but these systems are otherwise quite different form what is commonly considered as low molar mass liquid crystals .
historically , dipolar interactions where the first to be considered as the underlying interaction for the formation of mesophases : m. born , _ sitz .
_ * 25 * , 614 ( 1916 ) ; m. born , _ ann .
phys . _ * 55 * , 221 ( 1918 ) . j.w .
goodby , _ handbook of liquid crystals : low molecular weight liquid crystals i : calamitic liquid crystals , _ edited by d. demus , j.w .
goodby , g.w .
gray , h.w .
spiess and v. vill , ( wiley - vch , ny , 1998 ) , and references therein .
d. demus , j.w .
goodby , g.w .
gray , h.w .
spiess and v. vill ( editors ) , _ handbook of liquid crystals : low molecular weight liquid crystals ii : discotic and non - conventional liquid crystals _ ( wiley - vch , ny , 1998 ) , and references therein . | various possibilities of polar self - organisation in low molar mass nematic , smectic and columnar liquid crystals are discussed with particular focus on the underlying molecular symmetries and interactions .
distinction is made between vector and pseudovector polarities , their quantification in terms of molecular order parameters and their relation to spontaneous electric polarisation and to molecular chirality .
the understanding of the molecular mechanisms that give rise to polar ordering in existing lamellar and columnar phases may be useful for the design of new polar variants of common a - polar liquid crystals .
_ keywords : _ polar nematics ; ferroelectric liquid crystals ; polar ordering . |
the inflationary paradigm @xcite offers the attractive possibility of resolving many of the puzzles of standard hot big bang cosmology .
the crucial ingredient of most successful inflationary scenarios is a period of `` slow - roll '' evolution of a scalar field @xmath1 ( the `` inflaton '' ) , during which the potential energy @xmath2 stored in @xmath1 dominates its kinetic energy @xmath3 and drives a quasi - exponential expansion of the universe . at present
there exists no preferred concrete inflationary scenario based on a convincing realistic particle physics model .
in particular , though string theory provides one with several very weakly coupled scalar fields ( the moduli ) , which could be natural inflaton candidates , their non - perturbative potentials @xmath2 do not seem fit to sustain slow - roll inflation because , for large values of @xmath1 , they tend either to grow , or to tend to zero , too fast .
it is therefore important to explore novel possibilities for implementing an inflationary evolution of the early universe .
the aim of this work is to point out that , even in absence of any potential energy term , a general class of non - standard ( i.e. non - quadratic ) _ kinetic - energy _ terms @xmath4 , for a scalar field @xmath1 , can drive an inflationary evolution of the same type as the usually considered potential driven inflation . by `` usual type of inflation ''
we mean here an accelerated _ expansion _
( in the einstein conformal frame ) during which the curvature scale starts around a planckian value and then _ decreases _ monotonously .
by contrast , the pre - big bang scenario @xcite uses a _ standard _ ( quadratic ) kinetic - energy term @xmath3 to drive an accelerated _ contraction _ ( in the einstein frame ) during which the curvature scale _
increases_. though we shall motivate below the consideration of non - standard kinetic terms by appealing to the existence , in string theory , of higher - order corrections to the effective action for @xmath1 , we do not claim that the structure needed for implementing our `` kinetically driven '' inflation arises inevitably in string theory .
the aim of this work is more modest .
we draw attention to a new mechanism for implementing inflation . a large class of the ( toy ) models we shall consider satisfy two of the most crucial requirements of inflationary scenarios : ( i ) the scalar perturbations are `` well - behaved '' during the inflationary stage , and ( ii ) there exist natural mechanisms for exiting inflation in a `` graceful '' manner .
we leave to future work a more detailed analysis of the observational consequences of our kinetically driven inflation ( spectrum of scalar and tensor perturbations , reheating , ) .
we hope that , by enlarging the basic `` tool - kit '' of inflationary cosmology , our mechanism could help to locate which sector of string - theory ( if any ) has inflated a strongly curved initial state into our presently observed large and weakly curved universe .
we consider a single scalar field @xmath1 interacting with gravity through non - standard kinetic terms , @xmath5 .\ ] ] here , @xmath6 and we use the signature @xmath7 . for simplicity ,
we only consider lagrangians @xmath8 which are local functions of @xmath9 and therefore , depend only on the scalar @xmath10 as we bar here the consideration of potential terms , we must impose that the function @xmath11 vanish when @xmath12 . near @xmath13 , a generic kinetic lagrangian @xmath11 is expected to admit an expansion of the form @xmath14 one of the possible particle theory motivations for taking this kind of lagrangian seriously could be the following .
let us consider only gravity and some moduli field @xmath1 ( which could be the dilaton , or some other moduli ) in string theory .
it is well - known that @xmath15 corrections ( due to the massive modes of the string ) generate a series of higher - derivative terms in the low - energy effective action @xmath16 , while string - loop corrections generate a non - trivial moduli dependence of the coefficients of the various kinetic terms .
this leads to a structure of the type ( in the string frame @xmath17 ) @xmath18 + o(\alpha'\,^{2})\big\ } , \nonumber\end{aligned}\ ] ] where the ellipsis stands for other four - derivative terms ( like @xmath19 . in the case where @xmath1 is the dilaton the coupling functions are of the form @xmath20 where the ellipsis contain higher contributions in @xmath21 , including non - perturbative ones . transforming eq .
( [ eq : string - effective - action ] ) to the einstein frame @xmath22 , and neglecting the other possible four derivative terms in eq .
( [ eq : string - effective - action ] ) , leads to the effective action ( [ eq : einstein - action ] ) with @xmath23-kinetic terms of the form ( [ eq : p - expansion ] ) where @xmath24 in the case where @xmath1 is the dilaton ( so that @xmath25 is the string coupling ) , we have , in the weak - coupling limit @xmath26 , @xmath27 so that @xmath28 and @xmath29 .
however , when @xmath30 becomes of the order of unity it is not a priori excluded that @xmath31 and @xmath32 could become more complicated functions of @xmath23 .
we shall give below some examples of such possible complicated behaviours .
returning now to a general lagrangian @xmath33 the `` matter '' energy - momentum tensor reads @xmath34 equation ( [ eq : energy - momentum - tensor ] ) shows that , if @xmath35 is time - like ( i.e. @xmath36 ) , our scalar field action is `` equivalent '' to a perfect fluid ( @xmath37 ) with pressure @xmath38 energy density @xmath39 and four - velocity @xmath40 where @xmath41 denotes the sign of @xmath42 . as usual , in inflationary cosmology
, we consider a flat background friedmann model @xmath43 and an homogeneous background scalar field @xmath44 .
a convenient minimal set of independent evolution equations for @xmath45 and @xmath46 is ( with @xmath47 ) @xmath48 it will be also useful to refer to other ( redundant ) forms of the evolution equations : @xmath49 where @xmath50 denotes the momentum conjugate to @xmath23 .
[ note that equation ( [ eq : energy - density ] ) reads @xmath51 , which is the usual energy associated to the lagrangian @xmath52 . ] in this work we shall only consider solutions of eqs .
( [ eq : friedmann])-([eq : canonical ] ) which describe an expanding universe ( in einstein frame ) , that is @xmath53 this reduces the evolution equations ( [ eq : friedmann ] ) , ( [ eq : energy - conservation ] ) to the master equation @xmath54 here , and in the following , we use units such that @xmath55 .
note that , from eq .
( [ eq : friedmann ] ) , @xmath56 was constrained to be positive .
as a warm up , let us first consider the case where the lagrangian @xmath8 depends only on @xmath57 , and not on @xmath1 : @xmath58 . from eq .
( [ eq : energy - density ] ) , the energy - density depends also only on @xmath59 : @xmath60 . in hydrodynamical language
it means that we have here an `` isentropic '' fluid , with a general equation of state relating @xmath8 to @xmath56 : @xmath61 .
the master evolution equation can then be qualitatively solved by looking at the graph of the equation of state @xmath61 .
the shape of this graph depends very much on the shape of the function @xmath58 , or better , on the shape of the function @xmath62 where @xmath63 , so that @xmath64 .
indeed , the relation @xmath65 shows that @xmath56 can be read geometrically off the graph @xmath62 .
more precisely @xmath66 is the `` intercept''of the tangent to the curve @xmath62 at the point @xmath67 , i.e. its intersection with the vertical axis @xmath68 .
the minus sign means that @xmath56 is positive ( negative ) if the tangent intersects the vertical axis below ( above ) the @xmath69 horizontal axis .
therefore , if the function @xmath70 is , for instance , always _ convex _ , @xmath71 , as will be the case if all the coefficients @xmath72 appearing in the expansion ( [ eq : p - expansion ] ) are positive , @xmath8 and @xmath56 will be always positive .
in such a case , eq .
( [ eq : master ] ) shows that @xmath56 will monotonically decrease towards zero , and the evolution will be driven to the `` attracting solution '' @xmath73 this attractor corresponds to the asymptotic equation of state @xmath74 valid near @xmath75 where the usual kinetic term @xmath76 dominates . on the other hand , if the function @xmath77 is non - convex and has some oscillatory behaviour as @xmath78 increases ( i.e. , if we consider the general case where the expansion coefficients @xmath72 in eq .
( [ eq : p - expansion ] ) may take negative values ) the graph @xmath61 can be more complicated and can allow for exponential - type inflationary behaviour .
let us first note that , because of eq .
( [ eq : varepsilon - psi ] ) , the extrema of the function @xmath79 ( or @xmath58 ) correspond to values where @xmath80 , i.e. to _ fixed points _ of the master evolution equation ( [ eq : master ] ) . for a general function @xmath58
the graph of the ( multiform ) equation of state might resemble fig.[fig : state ] . from the master eq .
( [ eq : master ] ) it follows that @xmath56 will decrease above the line @xmath80 , and increase below it .
[ fig : state ] then shows that all the intersection points with the @xmath80 line are _ attractors _ of the ( future ) evolution .
the arrows in fig .
[ fig : state ] indicate the evolutive flow , which reverses ( along the graph ) at the extrema of @xmath56 .
the region where @xmath81 is excluded because it can not be reached by flat cosmologies ( see eq . (
[ eq : friedmann ] ) ) .
the fixed points lying at the @xmath80 line correspond to an exponential inflation @xmath82 apart from these inflationary attractors , there are two other attractors ( in the case depicted in fig .
[ fig : state ] ) : ( i ) the origin , where the evolution is driven toward the solution ( [ eq : vacuum - attractor ] ) ( corresponding to the `` hard '' equation of state @xmath83 ) , and ( ii ) the point above the origin on the vertical axis . as we will see
later the latter point lies in the region of absolute instability and has therefore no physical significance . in this work
we shall focus on the inflationary attractors ( [ eq : inflationary - attractor ] ) because they exhibit the novel possibility of getting , for a large set of initial conditions , quasi - exponential ( or power law ) inflation out of a purely kinetic lagrangian .
note that the condition for the existence of these inflationary attractors can also be seen in eq .
( [ eq : canonical ] ) . in absence of a @xmath1 dependence , eq .
( [ eq : canonical ] ) says that @xmath84 is constant so that @xmath85 is attracted toward zero . as the momentum @xmath86
, we see that non trivial ( @xmath87 ) attractors can exist if the kinetic terms are non - standard so that @xmath88 can vanish for non zero values of @xmath89 .
the extremal values of @xmath8 correspond to the inflationary attractors discussed above .
the labels `` s '' and `` u '' in fig .
[ fig : state ] ( which stand for `` stable '' and `` unstable '' ) indicate whether , for our present isentropic equation of state @xmath90 , the squared speed of sound @xmath91 is positive or negative , respectively .
this issue will be further discussed below .
note that the `` price '' to pay for having inflationary attractors as in fig .
[ fig : state ] is the existence of regions , in phase space @xmath92 , with negative energy density .
we shall assume in this work that this is not physically forbidden .
all the cosmological evolutions that we shall consider below stay always in the positive energy regions , and the existence , elsewhere in phase space , of negative @xmath56 regions does not necessarily cause some instabilities along our evolutionary tracks .
the simple case of a kinetic lagrangian depending only on @xmath89 considered above is the analog , for kinetically driven inflation ( or `` _
_ k__-inflation '' for short ) , of a de sitter model with constant energy density .
it is clear that both models should have similar problems .
namely , there is no natural graceful exit , no smooth transition to a friedmann universe and the cosmological perturbations are `` ill - defined '' .
to avoid these problems we should allow the coefficients in the expansion ( [ eq : p - expansion ] ) of the lagrangian @xmath8 to depend on the scalar field @xmath23 .
the simplest way to realize successfully the idea of _ k_-inflation is to consider the analog of `` slow - roll '' potential driven inflation , in which the potential @xmath2 in the lagrangian @xmath93 dominates the kinetic term @xmath3 and evolves slowly . for the concept of _ k_-inflation to have a relevance to a large class of models , we need to consider a general kinetic lagrangian @xmath33
the idea is therefore to find the conditions under which the influence of the non - trivial @xmath1 dependence of @xmath33 will represent only a relatively small perturbation of the attraction toward exponential inflation discussed in section [ sec : k - inflation ] . to do that in a concrete manner
it is convenient to focus henceforth on the simplest kinetic lagrangian , containing only @xmath94 and @xmath95 terms , namely @xmath96 let us first motivate the possibility of rather arbitrary functions @xmath97 by considering again the low - energy effective action of string theory ( [ eq : string - effective - action ] ) . as we mentioned above , in the weak coupling limit @xmath98 and @xmath99 however ,
when @xmath100 becomes of order unity it is not a priori excluded that @xmath101 could change sign .
for instance we could consider a simple model ( of the type considered in refs .
@xcite , @xcite ) , where the coupling functions @xmath102 in the action ( [ eq : string - effective - action ] ) are the same , that is @xmath103 . in this model
@xmath104 might become negative when @xmath100 reaches values of order unity .
in fact a model , incorporating string loop corrections , considered in ref .
@xcite has @xmath105 where @xmath106 is a positive parameter .
the r.h.s . of eq .
( [ eq : fms - k ] ) becomes negative when @xmath107 .
motivated by these examples , we shall assume , as simplest toy model exhibiting interesting dynamics , a lagrangian of the form ( [ eq : simplest - p - expansion ] ) with a function @xmath108 which is positive in some range of values of @xmath1 ( `` weak - coupling - domain '' ) and becomes negative in some other range ( `` strong - coupling - domain '' ) . on the other hand , to ensure the positivity of @xmath56 for large field gradients @xmath59 , we shall assume that the function @xmath109 remains always positive .
the equation of state in the model ( [ eq : simplest - p - expansion ] ) is parametrically given by @xmath110 we represent in fig .
[ fig : model - state ] the change in the form of the equation of state @xmath111 as @xmath23 varies from the weak - coupling region ( @xmath112 ) to the strong - coupling one ( @xmath113 ) . note that for large values of @xmath59 , the equation of state asymptotes the one of radiation ( @xmath114 ) , while it is tangent to the hard equation of state ( @xmath115 ; with @xmath116 ) when @xmath117 . in the strong coupling domain there appears ( in the adiabatic approximation where @xmath23 is treated as constant ) an inflationary fixed point where @xmath118 .
let us investigate under what conditions on the functions @xmath101 and @xmath32 one can indeed approximately solve the master evolution equation ( [ eq : master ] ) by considering that the variation of @xmath23 brings only a small perturbation onto the simple @xmath1-independent evolution studied above .
we shall simplify this study by redefining the scalar field and working with the new field variable @xmath119 ( which is well defined because we assume @xmath120 ) . with this definition @xmath121 and @xmath122 .
in other words , we can assume , without loss of generality , that @xmath123 . with this simplification the zeroth - order `` slow - roll '' , or `` adiabatic '' , solution to eq .
( [ eq : master ] ) , i.e. the instantaneous attractive fixed point of eq .
( [ eq : master ] ) ( solution of @xmath124 corresponds to @xmath125 here @xmath126 is positive ( in the slow - roll domain ) and @xmath41 denotes the sign of @xmath89 .
the time evolution of the slow - roll _ k_-inflation is given , from eqs .
( [ eq : x - slow - roll])-([eq : h - slow - roll ] ) , by simple quadratures ( the subscript `` in '' denotes initial values ) @xmath127 [ the notation @xmath128 is introduced to denote the number of @xmath129-folds of inflation . ] the post - slow - roll approximation , @xmath130 , is then obtained by rewriting the master equation ( [ eq : master ] ) as @xmath131 and replacing the slow - roll approximations ( [ eq : x - slow - roll])-([eq : h - slow - roll ] ) in the r.h.s .
this yields @xmath132 where the prime denotes a derivative with respect to @xmath1 .
the criterion for the validity of our previous slow - roll solution ( [ eq : x - slow - roll])-([eq : h - slow - roll ] ) is @xmath133 i.e. @xmath134 ( when keeping @xmath135 it would read @xmath136 @xmath137 ) .
this condition is as easily satisfied as the usual slow - roll condition for potential driven inflation .
examples of functions @xmath138 that satisfy this condition are : ( i ) any power law or exponential ( or super - exponential ) growth as @xmath139 , ( ii ) any levelling off of @xmath138 ( @xmath140 , with @xmath141 ) as @xmath142 , or ( iii ) a sufficiently fast pole - like growth of @xmath143 , with @xmath144 , as @xmath145 .
note that during slow - roll _
k_-inflation the following useful relation @xmath146 is satisfied , that is , the fractional compensation of the energy density by the negative pressure is proportional to the small parameter @xmath147 therefore , in those models where @xmath148 or equivalently , the energy density decreases in the course of expansion , @xmath149 is positive .
it is also obvious from ( [ eq : e - plus - p ] ) , that inflation ends when @xmath150 becomes of the order of unity , i.e. when the slow - roll condition ( [ eq : slow - roll - condition ] ) is violated . for any function @xmath138 ( and more generally for any @xmath151 ) satisfying the slow - roll criterion we can visualize our _ k_-inflationary behaviour as being one point ( i.e. one value of @xmath152 ) on an adiabatically varying equation of state graph of the type of figs .
[ fig : state ] or [ fig : model - state ] .
[ the adiabatic variation we mention corresponding to the fact that each graph corresponds to some specific , instantaneous value of @xmath23 , which is itself evolving ] .
( [ eq : slow - roll - parameter ] ) ( or its generalization to a generic @xmath11 ) then tells us that the point @xmath152 is always _ displaced _ away from the intersections of the ( @xmath1-instantaneous ) graph with the @xmath153 line .
overall , the qualitative behaviour of the solutions we are focussing on is the following : initially , we start with some representative point in the @xmath154 plane lying on an equation of state graph corresponding to some initial value of @xmath1 , deep into the strong coupling domain .
we assume that , for strong coupling , the slow - roll criterion is very well satisfied . in a first evolution stage , we can neglect the @xmath23-dependence of the equation of state because there is a fast attraction taking just a few @xmath129-folds of the representative point toward the nearest inflationary attractor ( this stage is described by the arrows in fig .
[ fig : state ] ) . after this initial stage
, we can consider that our representative point follows the ( post- ) slow - roll motion @xmath152 , corresponding to a representative point near but away from the @xmath80 line ( such a point is indicated in fig .
[ fig : model - state ] ) .
as the evolution continues , the slow - roll condition is less and less well satisfied and the representative point straggles more and more away from the @xmath80 line .
at some point in the evolution the slow - roll criterion ( [ eq : slow - roll - condition ] ) becomes violated ( @xmath155 ) and one naturally exits the inflationary stage .
we shall come back later to this exit mechanism
. the qualitative picture of the evolution just represented ( based on a succession of graphs in the @xmath156 plane ) can also be globally visualized by a phase - space picture in the @xmath157 plane ( see fig .
[ fig : phase ] ) .
`` slow - roll '' inflation on this graph corresponds to the portion of the separatrix ( attractor ) given by @xmath158 where @xmath159 when @xmath160 becomes of the order of one a graceful exit from inflation takes place , see section [ sec : exit - mechanisms ] .
the phase diagram fig .
[ fig : phase ] is very similar to the one of potential driven slow - roll inflation @xcite .
we see that the set of initial configurations of the scalar field which lead to inflation has nonzero measure .
therefore the problem of initial conditions here is very similar to what one has in the case of chaotic inflation @xcite .
we expect that in analogy to the other models of inflation self - reproduction of the universe @xcite can take place in our _ _ k-__inflationary model .
however this question needs a special investigation and we leave it to future work .
let us also mention that one can easily build a model in which one starts initially in the weak coupling regime , with the field @xmath23 evolving towards the strong coupling regime .
the function @xmath161 can then be arranged in such a way that the universe leaves the weak coupling regime to enter an inflationary stage and finally exits inflation .
it is well known that , in the usual potential driven inflationary scenario , if the potential depends exponentially on the scalar field , there exists an attractor solution which describes a power law inflating universe ( see for instance @xcite ) .
there is no graceful exit from inflation if the potential is exponential everywhere .
therefore to solve the graceful exit problem one should assume that the exponential potential is a valid approximation of a more realistic and complicated potential only within some limited range of values of the scalar field @xmath23 . as we show in this section , one can get an analogous power law _
k_-inflation within the class of models which we consider in this paper .
let us again consider the model with the lagrangian ( [ eq : simplest - p - expansion ] ) . for the purposes of this section
it is convenient to make a new field redefinition ( valid only in the region where @xmath113 ) and rewrite the lagrangian ( [ eq : simplest - p - expansion ] ) in terms of the new field variable @xmath162 where @xmath163 and @xmath164 . working with this lagrangian
one can try to find out whether there is a function @xmath165 for which the master equation ( [ eq : master ] ) has an _ exact _ solution which describes power law inflation . in the case of power law inflation @xmath166 where @xmath167 is a constant .
substituting @xmath56 and @xmath8 into the last equation we find immediately that if a solution exists , then @xmath168 expressing @xmath169 in terms of @xmath8 from ( [ eq : gamma ] ) and substituting ( [ eq : new - lagrangian ] ) with @xmath59 given by ( [ eq : power - law - x ] ) into the master eq .
( [ eq : master ] ) , we get a simple equation for @xmath170 , which is solved by @xmath171 therefore , a model with lagrangian ( [ eq : new - lagrangian ] ) , with @xmath165 given by eq .
( [ eq : power - law - f ] ) , has an attractor solution which describes power law expansion , @xmath172 if @xmath173 then this solution describes the usual power law inflation .
if one takes a negative value of @xmath167 then one gets pole - like super - inflation in _
einstein frame_. however , this pole - like inflation has a `` graceful exit problem '' which is very similar to the one of the pre - big bang scenario @xcite .
we were unable to find a simple solution to this problem and we doubt that such a solution can be meaningfully discussed within the effective field lagrangian formalism considered here .
therefore we think that pole - like inflation does not help toward bringing a solution of the main cosmological problems . in distinction from pole - like inflation , in the model of power law inflation the graceful exit problem can be easily solved if in some range of @xmath23 the function @xmath170 is modified in an obvious way .
a natural generalization of the lagrangian ( [ eq : new - lagrangian ] ) , @xmath174 where @xmath175 is a rather arbitrary function of @xmath59 , opens the possibility to realize power law inflation in a wide class of theories . actually , taking the function @xmath170 to be @xmath176 where @xmath177 is a solution of the equation
@xmath178 one can easily verify that power law inflation ( [ eq : power - law - a ] ) is a solution of the corresponding theory .
a detailed study of the spectrum of quantum perturbations in our slow - roll _ k_-inflation scenario will be done in a forthcoming publication .
we shall only discuss here a necessary condition for the _ stability _ of our models with respect to arbitrary , high - frequency scalar perturbations .
the equation for the canonical `` quantization variable '' @xmath179 describing the collective metric and scalar field perturbations in the case of the action ( [ eq : einstein - action ] ) can be written down in the standard way @xcite and takes the form @xcite @xmath180 where the only relevant piece of information for our stability analysis is the appearance of the `` speed of sound '' @xmath181 in front of the laplacian . here
a comma denotes a partial derivative with respect to @xmath182 it is clear that if @xmath183 is negative then the model is absolutely unstable .
the increment of instability is inversely proportional to the wavelength of the perturbations , and therefore the background models for which @xmath184 are violently unstable and do not have any physical significance .
this stability requirement ( @xmath185 ) is non trivial within our scenario , because , for instance , in slow - roll _ k_-inflation in the zeroth - order slow - roll approximation the inflationary attractors are defined by @xmath186 , and therefore @xmath187 .
however , as discussed in section [ sec : slow - roll - inflation ] , in the post - slow - roll approximation , with @xmath152 , @xmath188 does not vanish . to first order in @xmath149 we can write @xmath189 . using the second equation ( [ eq : sound - velocities ] ) we get @xmath190 therefore stability
requires that @xmath191 , i.e. that on the equation of state graphs of figs .
[ fig : state ] and [ fig : model - state ] , the @xmath1-gradients of @xmath11 be such that they displace the real , non - adiabatic , slow - roll attractor _ beyond _
the @xmath80 line ( `` beyond '' meaning here `` further away '' as one runs along the @xmath192 graph following the natural @xmath59 parametrization ) .
these stable stretches of the @xmath154 graphs are labelled @xmath193 in fig .
[ fig : state ] .
they are also the stretches where the slope @xmath194 is positive ( as is clear from eq .
( [ eq : sound - velocities ] ) which says that the velocity of sound is given by the usual formula @xmath195 , when keeping @xmath23 fixed ) .
let us now discuss the simple model ( [ eq : simplest - p - expansion ] ) .
for this model we have computed @xmath150 , and we can therefore assess under what conditions _
k_-inflation will be stable under scalar perturbations .
we see from eq .
( [ eq : slow - roll - parameter ] ) that the conditions can be expressed in two ( equivalent ) forms : ( i ) the energy density @xmath196 must _ decrease _ during the slow - roll of @xmath23 , or ( ii ) @xmath197 must be positive .
the satisfaction of this condition is very natural within the intuitive picture we have in mind : namely , starting at some high ( @xmath198 planckian ) energy density , i.e. a large _ negative _ value of @xmath199 , and then letting @xmath23 evolve toward the weak - field coupling domain where @xmath108 vanishes before becoming positive . during the slow - roll phase ( with @xmath200 ) , it is natural ( and even necessary if @xmath108 is monotonic ) to have a decreasing @xmath201 .
one consequence applicable to a general model @xmath11 is that slow - roll implies a small value @xmath202 .
it is interesting to ask for which ( non slow - roll ) models one can have both a continued _ k_-inflation and a constant speed of sound of order unity .
let us consider for that purpose power law inflation . in the case where the lagrangian takes the form ( [ eq : new - lagrangian ] ) the speed of sound during the inflationary stage is @xmath203
if we restrict ourselves to the inflationary range @xmath204 this speed can not exceed @xmath205 the smaller values of @xmath206 correspond to very fast ( nearly exponential ) expansion and small speed of sound in complete agreement with our analysis of slow - roll _ k_-inflation .
pole - like inflation is violently unstable in this model .
however , if we consider more general lagrangians ( [ eq : general - lagrangian ] ) we can avoid these restrictions .
actually in this case the speed of sound during the inflationary stage is given by the expression @xmath207 where @xmath177 is the solution of equation ( [ eq : solx ] ) .
the necessary conditions for power law inflation given in section [ sec : power - law - kinflation ] imply that @xmath208 and do not involve any restrictions on the second derivative @xmath209 .
therefore , for a power law inflationary stage with any a priori given value of the parameter @xmath167 , one can always find a corresponding function @xmath210 to arrange any required speed of sound .
note that it follows from here that one can also easily build a theory with pole - like inflation which is stable with respect to scalar perturbations .
in the simple model of section [ sec : slow - roll - inflation ] ( and in the normalization @xmath123 ) the total number of inflationary @xmath129-folds is ( considering for definiteness that @xmath23 decreases during slow - roll ) @xmath211 here @xmath212 is the initial value of @xmath23 , and @xmath213 the end of slow - roll , i.e. the value of @xmath23 where @xmath160 becomes of order unity , i.e. ( from eq .
( [ eq : slow - roll - parameter ] ) ) , such that @xmath214 we shall not investigate here the problem of the choice of initial conditions within our model .
we shall assume that some large parameter is present ( or at least possible ) in the problem and allows @xmath215 to be larger that 60 or so .
[ one simple possibility would be a function @xmath216 of the type of eq .
( [ eq : fms - k ] ) which levels off to a negative constant when @xmath1 increases .
all the couples @xmath217 where @xmath218 ( in planck units ) and @xmath1 is arbitrary large lead to an energy density of order @xmath219 . a random initial condition with @xmath220 could have an arbitrary large @xmath23 . ] assuming this we note not only that our mechanism contains a natural exit from inflation ( because of the evolution of @xmath161 and its final change of sign ) , but that this exit is generically expected to take place within a small number of @xmath129-folds .
indeed , condition ( [ eq : slow - roll - end ] ) signalling the end of slow - roll _ k_-inflation can be rewritten ( using eqs .
( [ eq : x - slow - roll])-([eq : h - slow - roll ] ) ) as @xmath221_{\rm end}\sim 1 $ ] , which means that @xmath161 changes by 100% in a hubble time , around the time of exit of _ k_-inflation .
one therefore expects that one can approximately match slow - roll _ k_-inflation with a post - inflationary phase where @xmath101 has become positive .
we think that , in most cases , this way of exiting _ k_-inflation provides a naturally graceful exit .
indeed , it is clear from fig . [
fig : model - state ] that , after the transition to the @xmath112 ( `` weak - coupling '' ) branch , the cosmological evolution will quickly be attracted toward an approximate @xmath222 equation of state . the corresponding expansion was discussed in eq .
( [ eq : vacuum - attractor ] ) and corresponds to a very fast decrease of the energy density : @xmath223 .
as this decrease is much faster than the decay of the energy density in radiation ( @xmath224 ) [ and in non relativistic matter ( @xmath225 ) if any is present ] , even small traces of the latter forms of energy present at the end of _
k_-inflation will ultimately dominate the expansion .
the situation is very similar to what has been recently discussed in ref . @xcite .
as we have in mind that , in our model , the scalar @xmath23 could be the dilaton ( or a moduli ) , i.e. a field which modifies the coupling constants of all the other matter fields , we expect that the nearly uniform time variation of @xmath23 during _ k_-inflation will generate quantum particles at a uniform spacetime rate . during _
k_-inflation the produced particles are constantly diluted by the fast expansion and are not expected to cause a strong back reaction , but the particles produced in the last @xmath129-fold of _
k_-inflation should be sufficiently numerous to dominate soon the expansion . even without this assumption ( that the couplings of @xmath23 are efficient in producing particles )
, the mere effect of the variable gravitational coupling ( at the end of inflation ) is sufficient to create any scalar particle with energy density ( at birth ) @xmath226 @xcite @xcite . to discuss more precisely the evolution of the inflaton
after it exits from _ k_-inflation , i.e. when @xmath108 has become positive , and the higher - derivative term @xmath227 has become negligible , it is convenient to introduce the canonical scalar field @xmath228 . in terms of @xmath229
the equation of motion reads @xmath230 , with @xmath231 .
hence @xmath232 and @xmath229 evolves according to @xmath233 . in a first phase after _
k_-inflation @xmath234 probably dominates over @xmath235 and the evolution follows the @xmath115 attractor solution eq .
( [ eq : vacuum - attractor ] ) . during this initial phase
@xmath236 so that @xmath237 drifts logarithmically : @xmath238 . later , @xmath239 will take over @xmath240 and the evolution will become radiation dominated . in this second phase ,
@xmath241 and @xmath242 stops drifting logarithmically to converge toward some final value @xmath243 : @xmath244 . note that , in this generic exit mechanism , the final value @xmath245 of the original field ( corresponding to the final value @xmath243 of the canonical field ) is arbitrary . therefore ,
if @xmath23 is the dilaton or a moduli our _ k_-inflationary mechanism does not , by itself , provide a mechanism for fixing @xmath23 to a particular value . to do that one must appeal either to the presence , at low - energies , of a @xmath23-dependent potential energy term @xmath246 ( which may have been negligible at high energies ) , or to a non - trivial structure of couplings to matter @xcite .
we wish , however , to point out that some variants of our general model can also provide another way of fixing the end location of @xmath23 very near a particular value .
indeed , if the kinetic function @xmath108 ( of the original @xmath23 variable ) happens to have a pole singularity @xmath247 with @xmath248 in the positive-@xmath161 domain the corresponding canonical field @xmath249 diverges when @xmath250 .
therefore , if this pole singularity is in the way of the evolution of @xmath23 after the exit from _ k_-inflation ( e.g. , if @xmath251 in the case where @xmath252 ) , the typically large logarithmic drift of the canonical field @xmath229 during the first phase after exit , @xmath253 ( where @xmath254 denotes the beginning of radiation domination ) , will mean that the original field @xmath23 will end up very near @xmath255 . evidently , this mechanism assumes that , except for @xmath108 , all the functions describing the coupling of @xmath23 to matter ( like the one giving the @xmath23-dependence of the gauge couplings ) are regular ( or , at least , less singular ) at @xmath256 .
then , in the notation of ref .
@xcite , the observable coupling strength of @xmath23 to matter @xmath257 is driven near zero by our mechanism .
we have pointed out that a general class of higher - order scalar kinetic terms @xmath258 can drive an inflationary evolution starting from rather arbitrary initial conditions . under not very restrictive conditions on the @xmath1-dependence of the kinetic term @xmath259 the early cosmological evolution
will be attracted toward a slow - roll kinetically driven inflationary stage . in a large class of models
this slow - roll behaviour has the following attractive features : ( i ) it drives the evolution from an initial high - curvature phase down to lower curvatures while dilating space in a quasi - exponential or power law manner , ( ii ) it is stable under ( high - frequency ) scalar perturbations , and ( iii ) it contains a natural exit mechanism because of the @xmath23-dependence of the kinetic terms .
we have briefly discussed the exit of kinetically driven inflation and found that it seems to be naturally `` graceful '' in lending itself to a smooth transition toward a stage dominated by the radiation produced ( either through the @xmath23-dependence , or through purely gravitational effects ) at the end of slow - roll .
we have also pointed out that the presence of pole - like singularities in the @xmath23-dependence of the kinetic terms can have some useful consequences : ( i ) a @xmath23-dependence of the form @xmath260 in the initial field domain can ensure a nearly constant speed of sound of order unity , while ( ii ) a pole - like singularity in the coefficient @xmath261 in the final field domain can help to fix the value of @xmath23 to a specific value .
we leave to future work the investigation of general important issues : the choice of ( a measure on the ) initial conditions , and the computation of the perturbation spectra generated by this new type of inflation .
we are aware that the compatibility of the latter issues with observations will probably necessitate the presence of some small ( or large ) dimensionless parameters in @xmath259 . in this work
, we have used the structure of the effective action in string theory as a partial motivation for considering higher - order kinetic terms @xmath259 .
having found that such terms can generically drive an inflationary behaviour , we hope that our mechanism might be useful in suggesting new ways in which the dilaton and moduli fields of string theory might be compatible with inflation .
the work of v.m . and c.a.p .
was supported in part by the _ sonderforschungsbereich 375 - 95 der deutschen forschungsgemeinschaft_. t.d.s work was partly supported by the nasa grant nas8 - 39225 to gravity probe b. c.a.p .
thanks the _ institut des hautes tudes scientifiques _ for its kind hospitality .
g. veneziano , phys . lett . *
b265 * , 287 ( 1991)m . gasperini and g. veneziano , astropart . phys . * 1 * , 317 ( 1993 ) for an updated collection of papers on the pre - big bang scenario see http://www.to.infn.it/gasperin/ | it is shown that a large class of higher - order ( i.e. non - quadratic ) scalar kinetic terms can , without the help of potential terms , drive an inflationary evolution starting from rather generic initial conditions . in many models ,
this kinetically driven inflation ( or `` @xmath0-inflation '' for short ) rolls slowly from a high - curvature initial phase , down to a low - curvature phase and can exit inflation to end up being radiation - dominated , in a naturally graceful manner .
we hope that this novel inflation mechanism might be useful in suggesting new ways of reconciling the string dilaton with inflation . _
k_-inflation + ludwig - maximilians - universitt , sektion physik + mnchen , germany + http://www.muk.physik.uni-muenchen.de + institut des hautes tudes scientifiques + bures sur yvette , france + http://www.ihes.fr + gravity probe b , hansen experimental physics laboratory + stanford university , stanford , u.s.a .
+ http://einstein.stanford.edu |
the design of spectrum sharing and access mechanisms for cognitive radio networks ( crns ) has attracted much attention in the last few years .
the interest in crns is mainly attributed to their ability to enable efficient spectrum utilization and provide wireless networking solutions in scenarios where the un - licensed spectrum bands are heavily utilized .
furthermore , due to their cognitive nature , crns are more spectrum efficient and robust than their non - cognitive counterparts against spectrum unavailability , and have the capability to utilize different frequency bands and adapt their operating parameters based on the surrounding radio frequency ( rf ) environment .
specifically , cr is considered as the key technology to effectively address the inefficient spectrum utilization in legacy licensed wireless communication systems by providing opportunistic on - demand access @xcite .
cr technology enables unlicensed users to opportunistically utilize the idle pr channels ( so - called spectrum holes ) .
the spectrum holes represent the pr channels that are currently under - utilized . in order to utilize these spectrum opportunities without interfering with the prns
, cr users should perform accurate spectrum sensing , through which idle channel lists are identified .
in addition , the cr users should be flexible enough to quickly vacate the operating channel when a pr user reclaims it . in this case
, cr users should quickly and seamlessly switch their operating channel(s ) .
while large - scale deployment of crns is still to come , extensive research attempts are currently underway to improve the effectiveness of spectrum sharing protocols and improve the spectrum management and operation of such networks @xcite .
two of the most crucial challenges in deploying crns are the needs to maximize spectrum efficiency and minimize the caused interference to prns . on other words
, providing efficient communication and spectrum access protocols that provide high throughput while protecting the performance of licensed prns is the crucial design challenge in crns .
the main objective of this paper is to overview and analyze the key schemes and protocols for spectrum access / sharing / managemnent that have been developed for crns in the literature .
furthermore , we briefly highlight a number of opportunistic spectrum sharing and management schemes and explain their operation details . as indicated later , it follows logically that cross - layer design , link quality / channel availability tradeoff and interference management are the key design principles for providing efficient spectrum utilization in crns .
we start by describing the main crn architectures and operating environment .
then , the spectrum sharing problem is stated .
the various objectives used to formulate the spectrum sharing problem in crns are summarized .
we then point out the several design challenges in designing efficient spectrum sharing and access mechanisms .
the tradeoffs in selecting the operating channel(s ) in crns are discussed .
a number of spectrum sharing design categories are then surveyed .
various complementary approaches , new technologies and optimization methods that have great potential in facilitating the design of efficient crn communication protocols are highlighted and discussed . finally , concluding remarks are provided with several open research challenges .
= 4.5 in = 1.5 in = 1.2 in
typical crn environment consists of a number of different types of prns and one or several crns .
the pr and cr networks geographically co - exist within the same area . in terms of network topology ,
two basic types of crns are proposed : centralized multi - cell crns and infrastructure - less ad hoc crns .
figure [ fig : crnc ] depicts a composition view of a crn operating environment consisting of an ad hoc crn and a multi - cell centralized crn that coexist with two different types of prns .
the different prns have license to transmit over orthogonal non - overlapping spectrum bands , each with a different licensed bandwidth .
pr users of a given prn operate over the same set of licensed channels .
cr users can opportunistically utilize the entire pr licensed and unlicensed spectrum . for ad
hoc multi - hop crns without centralized entity , it is necessary to provide distributed spectrum access protocols that allow each cr user to separately access and utilize the available spectrum . furthermore , for centralized multi - cell crns , it is desirable to provide ( 1 ) centralized spectrum allocation protocols that allocate the available channels to the different cr cells , and ( 2 ) centralized channel assignment mechanisms that enable efficient spectrum reuse inside each cell . in general , the channel availability model of each pr channel in a given locality is described by a two - state on / off markov process .
this model describes the evolution between idle ( off ) and busy ( on ) states ( i.e. , the on state of a pr channel indicates that the pr channel is busy , while the off state reveals that the pr channel is idle ) .
the model is further described by the stochastic distribution of the busy and idle periods , which are generally distributed .
the distributions of the idle and busy states depend on the pr activities .
we note here the on and off periods of a given channel are independent random variables . for a given channel @xmath0 , the average idle and busy periods are @xmath1 and @xmath2 , respectively . based on this model , the idle and busy probabilities of a pr channel @xmath0
are respectively given by @xmath3 and @xmath4 .
figure [ fig : onoff ] shows a transition diagram of a @xmath5-state busy / idle markov model of a given pr channel .
we note here that neighboring cr users typically have similar views to spectrum availabilities , while non - neighboring cr users have different channel availability conditions .
= 3.2 in = 2.0 in = 1.5 in = 1.2 in
spectrum sharing problem ( including spectrum management and decision ) can be stated as follows : given the output of spectrum sensing , the main goal is to determine which channel(s ) to use , at what powers , and at what rates ?
such that a given performance metric ( objective function ) is optimized .
this is often a joint optimization problem that is very difficult to solve ( often it constitutes an np - hard problem ) .
recently , several spectrum assignment strategies have been proposed for crns @xcite .
these strategies are designed to optimize a number of performance metrics including : * maximizing the cr throughput ( individual users or network - level ) based on shannon capacity or a realistic staircase rate - sinr function ( e.g. , @xcite ) . * minimizing the number of assigned channels for each cr transmission ( e.g. , @xcite ) . * maximizing the cr load balance over the different pr channels ( e.g. , @xcite ) . * minimizing the probability of pr disruption , i.e. , minimizing the pr outage probability ( e.g. , @xcite ) . * minimizing the average holding time of selected channels ( minimizing the pr disruption ) ( e.g. , @xcite ) . * minimizing the frequency of channel switching due to pr appearance by selecting the channel with maximum residual idle time , i.e. , minimizing the cr disruption in terms of forced - termination rate ( e.g. , @xcite ) . * maximizing the cr probability of success ( e.g. , @xcite ) . * minimizing the spectrum switching delay for cr users ( e.g. , @xcite ) . * minimizing the expected cr waiting time to access a pr channel ( e.g. , @xcite ) . * minimizing crn overhead and providing cr qos ( e.g. , @xcite ) . * minimizing the overall energy consumption ( e.g. , @xcite ) . * achieving fair spectrum allocation and throughput distribution in the crn ( e.g. , @xcite ) . * maintaining crn connectivity with predefined qos requirements ( e.g. , @xcite ) .
we note here that the spectrum sharing problem for any of the aforementioned objectives is , in general , np - hard .
therefore , several heuristics algorithms and approximations have been proposed to provide suboptimal solutions for the problem in polynomial - time .
these heuristics and approximations can be classified based on their adopted optimization method as : graph theory - based algorithms ( e.g. , @xcite ) , game theory - based algorithms ( e.g. , @xcite ) , genetic - based algorithms ( e.g. , @xcite ) , linear programming relaxation - based algorithms ( e.g. , @xcite ) , fuzzy logic - based algorithms ( e.g. , @xcite ) , dynamic programming - based algorithms ( e.g. , @xcite ) , sequential - fixing - based algorithms ( e.g. , @xcite ) .
the coexistence problem is one of the most limiting factors in achieving efficient cr communications . in a crn environment
, there are three kinds of harmful interference that should be considered : pr - to - cr / cr - to - pr interference ( the so - called pr coexistence ) and cr - to - cr interference ( the so - called self - coexistence ) . while several mechanisms have been proposed to effectively deal with the pr - to - cr interference problem based on cooperative ( e.g.,@xcite ) or noncooperative ( e.g.,@xcite ) spectrum sensing , the cr - to - cr and cr - to - pr interference problems are still challenging issues . to address the cr - to - cr interference problem in ad hoc crns ,
several channel allocation and self - coexistence management mechanisms have been proposed based on either ( 1 ) exclusive channel assignment or ( 2 ) joint channel assignment and power control . on the contrary
, the cr - to - cr interference problem has been addressed in multi - cell centralized crns based on either fixed channel allocation @xcite or adaptive traffic - aware / spectrum - aware channel allocation @xcite .
it has been shown that the cr - to - pr interference is the most crucial interference in crn environment , because it has a direct effect on the performance of prns .
hence , the transmission power of cr users over the pr channels should be adaptively computed such that the performance of the prns is protected .
based on the outcomes of spectrum sensing , two different power control strategies can be identified : binary and multi - level transmission power strategies . according to the binary - level strategy (
the most widely used power control strategy in crns ) , cr users can only transmit over idle channels with no pr activities .
specifically , for a given pr channel , a cr transmits @xmath6 power if the channel is busy , and uses the maximum possible power if the pr channel is idle .
while this strategy ensures collision - free spectrum sharing between the cr and pr users , it requires perfect spectrum sensing .
worse yet , the binary - level strategy can lead to non - optimal spectrum utilization . on the other hand , using a multi - level adaptive frequency - dependent transmission power strategy allows the cr and pr users to simultaneously share the available spectrum in the same locality , which can significantly improve spectrum utilization . by allowing cr users to utilize both idle and partially - occupied pr channels
, much better spectrum utilization can be achieved .
the multi - level power strategy can also be made time - dependent to capture the dynamics of pr activities . under this strategy ,
controlling the cr - to - pr interference is nontrivial .
in addition , computing the appropriate and adequate multi - level power strategy is still a challenging issue , which has been studied under some simplified assumptions .
specifically , the authors in @xcite proposed an adaptive multi - level frequency- and locality - dependent cr transmission power strategy that provides a soft guarantee on prns performance . this adaptive strategy is dynamically determined according to the pr traffic activities and interference margins .
in this section , we review several well - known distributed coordination mechanisms designed for crns .
we note that control channel designs for crns can be loosely classified into six different categories @xcite : * dedicated pre - specified out - of - band common control channel ( ccc ) design @xcite . *
non - dedicated pre - specified in - band ccc design @xcite
. * hybrid out - of - band and in - band ccc design @xcite .
* hopping - based control channel design @xcite .
* spread - spectrum - based control channel design @xcite . * cluster - based local ccc design @xcite .
despite the fact that using a dedicated out - of - band ccc is straightforward , it contradicts the opportunistic behavior of crns , may result in a single - point - of - failure ( spof ) and performance bottleneck due to ccc saturation under high cr traffic loads .
similarly , using a pre - specified non - dedicated in - band ccc is not a practical solution due to spectrum heterogeneity and , if exists , such solution can result in a spof , become a performance bottleneck , and introduce security challenges .
another approach that can effectively deal with the ccc saturation issue ( bottleneck problem ) is to use a hybrid out - of - band and in - band ccc ( simultaneous control communications over in - band pr channels and dedicated out - of - band cccs ) .
this approach exploits the strengths of out - of - band and in - band signaling and , hence , can significantly enhance the performance of multi - hop crns . using a hopping - based control channel
can address the spof , bottleneck and security issues .
however , in such type of solutions , the response to pr appearance is challenging as cr users can not use a pr channel once reclaimed by pr users . in addition , this type of solution is generally spectrum unaware .
another key design issue in such solutions is the communication delay that heavily depends on the time to rendezvous .
using cluster - based coordination solutions , where neighboring cr users are dynamically grouped into clusters and establish local cccs , can provide reliable distributed coordination in crns @xcite .
however , adopting this type of solutions in a multi - hop crn is limited by several challenges , such as providing reliable inter - cluster communication ( i.e. , different cluster may consider different cccs ) , maintaining connectivity , broadcasting control information , identifying the best / optimal cluster size , and maintaining time - synchronization @xcite .
finally , using spread - spectrum - based distributed coordination is a promising solution to most of the aforementioned design challenges , but the practicality and design issues of such solution need to be further investigated . according to this solution , the control information is spread over a huge pr bandwidth with a very low transmission power level ( below the noise level ) .
consequently , with a proper design , an efficient ccc design can be implemented using spread spectrum with minor effect on prns performance . in conclusion
, various distributed coordination mechanisms have been developed to provide reliable communications for crns , none of which are totally satisfactory . hence , designing efficient distributed coordination schemes in crns should be based on novel coordination mechanisms along with effective transmission technologies that enable effective , robust and efficient control message exchanges .
the spectrum ( channel ) assignment problem in crns has been extensively studied in the literature .
existing channel assignment / selection solutions can loosely be classified into three categories : best link - quality schemes , larger availability - period schemes , and joint link - quality and channel - availability - aware schemes .
it has been shown ( e.g.,@xcite ) that using the best link - quality schemes in crns , where the idle channel(s ) with the highest transmission rate(s ) are selected , can only provide good performance under relatively static pr activities with average pr channel idle durations that are much larger than the needed transmission times for cr users @xcite . under highly dynamic pr activities ,
this class of schemes can result in increasing the cr termination rate , leading to a reduction in crn performance as a cr user may transmit over a good - quality pr channel with relatively short availability time ( short channel - idle period ) . on the other hand , employing
the larger availability - period schemes in crns ( e.g. , @xcite ) can result in increasing the cr forced - termination rate as an idle pr channel of very poor link - quality ( low transmission rate ) may be chosen , resulting in a significant reduction in crn performance .
we note here that the interaction between the crn and prns is fundamental for conducting channel assignment in crns .
the above discussion presents sufficient motivation to jointly consider the link - quality and average idle durations of pr channels when assigning operating channels to cr users .
however , several open questions in this domain still need to be addressed ; possibly the most challenging one is how to jointly consider the link - quality and average idle durations into one metric to perform channel assignment .
other important questions are : how can a cr user estimate the distribution of the idle periods of the different pr channels ?
what are the implications of the interaction between the crn and the prns
? how can a cr user determine the link - quality conditions over the various ( large number ) pr channels ? some of these questions have been addressed in @xcite by introducing the cr packet success probability metric .
this metric is derived based on stochastic models of the time - varying pr behaviors .
the probability of success over a given channel is a function of both the link - quality condition and the average - idle period of that pr channel .
it has been proven that it is necessary to jointly consider the link - quality conditions and availability times of available pr channels to improve the overall network performance @xcite .
there are several attempts have been made to design spectrum sharing protocols with the objective of improving the overall spectrum utilization while protecting the performance of licensed prns .
existing spectrum sharing / access protocols and schemes for crns can loosely be categorized into four main classes based on : the number of radio transceivers per cr user ( single - transceiver , dual - transceiver , or multiple transceiver ) , their reaction to pr behavior ( reactive , proactive , or interference - based threshold ) , their spectrum allocation behavior ( exclusive or non - exclusive spectrum occupancy model ) , and the guardband considerations ( guardband - aware or guardband - unaware ) .
spectrum sharing protocols and schemes for crns can also be categorized based on the number of radio transceivers per a cr user ( i.e. , single transceiver @xcite , dual transceivers @xcite , and multiple transceivers @xcite ) . using multiple ( or dual ) transceivers
greatly simplifies the task of spectrum access design and significantly improve system performance .
this is because a cr user can simultaneously utilize multiple channels ( the potential benefits of utilizing multi - channel parallel transmission in crns were demonstrated in @xcite ) .
in addition , the spectrum access issues such as hidden / exposed terminals , transmitter deafness and connectivity can be easily overcome as one of the transceivers can be switched to the assigned control channel ( i.e. , cr users can always receive control packet over the ccc even when they are operating over the data channels ) .
however , the achieved performance gain of using multiple transceivers ( multi - channel parallel transmission ) comes at the expense of extra hardware .
worse yet , the optimal joint channel assignment and power control problem in multi - transceiver crns is , in general , np - hard . on the other hand , it has been shown that the design of efficient channel assignment schemes for single - transceiver single - channel low - cost crns is simpler than that of the multi - transceiver counterpart @xcite . while single - transceiver designs can greatly simplify the task of finding the optimal channel assignment , the aforementioned channel access issues are not trivial and the performance is limited to the capacity of the selected channel .
spectrum sharing schemes in the crns can also be classified based on their reaction to the appearance of pr users into three main groups : ( 1 ) proactive ( e.g. , @xcite ) , ( 2 ) reactive ( e.g. , @xcite ) , and ( 3 ) interference threshold - based ( e.g. , @xcite ) . in reactive schemes ,
the active cr users switch channels after the pr appearance .
on the other hand , in proactive schemes , the cr users predict the pr appearance and switch channels accordingly .
the threshold - based schemes allow the cr users to share the spectrum ( both idle and partially - occupied pr channels ) with pr users as long as the interference introduced at the pr users is within acceptable values .
existing threshold - based schemes attempt at reducing the impacts of the un - controllable frequency - dependent pr - to - cr interference on crn performance through proper power control based on either ( 1 ) the instantaneous sensed interference @xcite , ( 2 ) the average measured pr interference @xcite , or ( 3 ) using stochastic pr interference models @xcite .
the spectrum sharing model represents the type of interference model used to solve the channel and power assignment problem .
there are two different spectrum sharing models : protocol ( interference avoidance ) and physical ( interference ) models @xcite .
the former employs an exclusive channel occupancy strategy , which eliminates the cr - to - cr interference and simplifies the management of the cr - to - pr interference @xcite .
however , it does not support concurrent cr transmissions over the same channel , which may reduce the spectrum efficiency . on the other hand ,
the overlay physical model allows for multiple concurrent interference - limited cr transmissions to simultaneously proceed over the same channel in the same locality , which improves spectrum efficiency @xcite .
however , the power control issue ( cr - to - cr and cr - to - pr interference management ) under this model is not trivial .
worse yet , using this model requires a distributed iterative power adjustment for individual cr users , which was shown that it results in slow protocol convergence @xcite .
most of existing spectrum sharing protocols for crns were designed assuming orthogonal channels , where the adjacent channel interference ( aci ) is ignored ( e.g. , @xcite ) .
however , this requires using ideal sharp transmit and receive filters , which is practically not feasible . in practice ,
frequency separation ( guard bands ) between adjacent channels is needed to mitigate the effects of aci and protect the performance of ongoing pr and cr users operating over adjacent channels .
it has been shown that introducing guard bands can significantly impact the spectrum efficiency , and hence it is very important to account for the guard - band constraints when designing spectrum sharing protocols for crns .
few number of crn spectrum access and sharing protocols have been designed while accounting for the guard band issue @xcite .
guard band - aware strategies enable effective and safe spectrum sharing , have a great potential to enhance the spectral efficiency , and protect the receptions of the ongoing cr and pr transmissions over adjacent channels .
the need for guard band - aware spectrum sharing mechanisms and protocols was discussed in @xcite .
specifically , the authors , in@xcite , have investigated the aci problem and proposed guard - band - aware spectrum access / sharing protocols for crns .
the main objective of their proposed mechanism is to minimize the total number of reserved guard - band channels such that the overall spectrum utilization is maximized . in @xcite
, the authors showed that selecting the operating channels on per block ( group of adjacent channels ) basis instead of per channel basis ( unlike the work in @xcite ) provides better spectrum efficiency .
the work in @xcite attempts at selecting channels such that at most one guard band is introduced for each new cr transmission .
in @xcite , the authors proposed two guard - band spectrum sharing mechanisms for crns .
the first mechanism is a static single - stage channel assignment that is suitable for distributed multi - hop crns .
the second one is an adaptive two - stage channel assignment that is suitable for centralized crns .
the main objective of the proposed mechanisms is to maximize spectrum efficiency while providing soft guarantees on cr performance in terms of a pre - specified rate demand .
in this section , we discuss and explain several methods and optimizations that interact with spectrum sharing protocols to further improve spectrum utilization in crns . the resource virtualization concept has been extensively discussed in the literature , which refers to the process of creating a number of logical resources based on the set of all available physical resources .
this concept allows the users to utilize the logical resources in the same way they are using the physical resources .
this leads to a better utilization of the physical resources as virtualization allows more users to share the available physical resources .
in addition , virtualization introduces an additional layer of security as user s application can not directly control the physical resources .
the concept of virtualization was originally used in computer systems to better utilize the available physical resources ( e.g. , processors , memory , storage units , and network interfaces ) .
these resources are virtualized into separate sets of logical resources , each set of these virtual resources can be assigned to different users .
using system virtualization can achieve : ( 1 ) users isolation , ( 2 ) customized services , and ( 3 ) improved resource efficiency .
virtualization was also been introduced in wired networks by introducing the framework of virtual private networks ( vpns ) . recently
, several attempts have been made to implement the virtualization concept in wireless crns .
we note here that employing virtualization in crns is daunted by several challenges including : spectrum sharing , limited infrastructure , different geographical regions , self co - existence , pr co - existence , dynamic spectrum availability , spectrum heterogeneity , and users mobility @xcite . in @xcite , a single cell crn virtualization framework was introduced .
according to this framework , a network with one bs and @xmath7 physical radio nodes ( pns ) with varying sets of resources are considered .
the resources include the number of radio interfaces at each pn , the set of orthogonal idle channels at each pn , and the employed coding schemes .
each pn hosts a set of virtual nodes ( vns ) .
the vns located in the different pns can communicate with each other . to facilitate such communications , vns request resources from their hosting pns .
simulation results have demonstrated the effectiveness of using network visrtualization in improving network performance . in @xcite ,
the authors have proposed a virtualization framework for multi - channel multi - cell crns . in this work ,
a virtualization based semi - decentralized resource allocation mechanism for crns using the concept of multilayer hypervisors was proposed .
the main objective of this work is to reduce the overall cr control overhead by minimizing the cr users reliance on the base - station in assigning spectrum resources .
simulation results have indicated significant improvement in crn performance ( in terms of control overhead , spectrum utilization and blocking rate ) is achieved by the virtualized framework compared to non - virtualized resource allocation schemes .
the problem of computing the optimal spectrum access strategy for cr users has been well investigated in @xcite , but for cr users that are equipped with half - duplex ( hd ) transceivers .
it has been shown that using hd transceivers can significantly reduce the achieved network performance @xcite .
motivated by the recent advances in full - duplex ( fd ) communications and self - interference suppression ( sis ) techniques , several attempts have been made to exploit the fd capabilities and sis techniques in designing communication protocols for crns @xcite .
the main objective of these protocols is to improve the overall spectrum efficiency by allowing simultaneous transmission and reception ( over the same channel or over different channels ) at each cr user .
these protocols , however , require additional hardware support ( i.e. , duplexers ) .
the practical aspects of using fd radios in crns need to be further investigated .
the design of effective channel / power / rate assignment schemes for fd - based crns is still an open problem .
beamforming techniques are another optimization that can enable efficient spectrum sharing @xcite . according to beamforming ,
the transmit and receive beamforming coefficients are adaptively computed by each cr user such that the achieved cr throughput is maximized while minimizing the introduced interference at the cr and pr users .
furthermore , the performance gain achieved by using beamforming in crns can be significantly improved by allowing for adaptive adjustment of the allocated powers to the transmit beamforming weights @xcite .
the operation details of such an approach need to be further explored .
the use of variable channel widths through channel aggregation and bonding is another promising approach in improving spectral efficiency .
however , this approach has not given enough attention .
based on its demonstrated excellent performance ( compared to using fixed - bandwidth channels ) , variable channel widths has been chosen as an effective spectrum allocation mechanism in cellular mobile communication systems , including the recently deployed 4 g wireless systems .
thus , it is very important to use variable - bandwidth channels in crns .
more specifically , in crns , assigning variable bandwidth to different cr users can be achieved through channel bonding and aggregation .
this has a great potential in improving spectrum efficiency .
the use of variable bandwidth transmission in crns is not straightforward due to the dynamic time - variant behavior of pr activities and the hardware nature of most of existing cr devices @xcite , which make it very hard to control the channel bandwidth @xcite .
so far , most of cr systems have been designed with the assumption that each cr user is equipped with single or several radio transceivers .
using hardware radio transceivers can limit the number of possibly assigned channels to cr users and can not fully support variable - width channel assignment .
one possible approach to enable variable - width spectrum assignment and increase network throughput is to employ software defined radios ( sdrs ) .
the use of the sdrs enables the cr users to bond and/or aggregate any number of channels , thus enabling variable spectrum - width cr transmissions .
thus , sdrs support more efficient spectrum utilization , which significantly improves the overall crn performance and provides qos guarantees to cr users .
cross - layer design is essential for efficient operation of crns .
spectrum sharing protocols for crns should select the next - hop and the operating pr frequency channel(s ) using a cross - layer design that incorporates the network , mac , and physical layers .
a cross - layer routing metric called the maximum probability of success ( mpos ) was proposed in @xcite .
the mpos incorporates the link quality conditions and the average availability periods of pr users to improve the crn performance in terms of the network throughput .
the metric assigns operating channels to the candidate routes so that a route with the maximum probability of success and minimum cr forced termination - rate is selected .
the main drawback of the mpos approach is its requirement of known pr channel availability distributions ( the probability density function of idle periods of the pr channels ) .
based on the spectrum availability conditions and to enable efficient crn operation , a cr user may need to utilize multiple adjacent ( contiguous ) idle pr channels ( the so - called spectrum bonding ) or non - adjacent ( non - contiguous ) idle pr channels ( the so - called spectrum aggregation ) .
spectrum bonding and aggregation can be realized using either the traditional frequency division multiplexing ( fdm ) or the discontinuous - orthogonal frequency division multiplexing ( d - ofdm ) technology @xcite .
the former technology requires several half - duplex transceivers and tunable filters at each cr user , where each assigned channel will use one of the available transceivers .
while this approach is simple , it requires a large number of transceivers and does not provide the enough flexibility needed to implement channel aggregation and bonding at a large - scale .
the d - ofdm is a novel wireless radio technology that allows a cr transmission to simultaneously take place over several ( adjacent or non - adjacent ) channels using one half - duplex ofdm transceiver . according to d - ofdm ,
each channel includes a distinct equal - size group of adjacent ofdm sub - carriers . according to d - ofdm
, spectrum bonding and aggregation with any number of channels can be realized through power control , in which the sub - carriers of a non - assigned channel will be assigned @xmath6 power and all the sub - carries of a selected channel will be assigned controlled levels of powers .
we note here that the problem of assigning different powers to different ofdm symbols within the same channel is still an open issue . multiple input multiple output ( mimo )
is considered as a key technology to increase the achieved wireless performance .
specifically , mimo can be used to improve spectrum efficiency , throughput performance , wireless capacity , network connectivity , and energy efficiency .
the majority of previously proposed works on mimo - based crns ( e.g. , @xcite ) have focused on the physical layer and addressed a few of the challenging issues at the upper layers , but certainly more effort is still required to investigate the achieved capacity of mimo - based crns , the design of optimal channel / power / rate assignment for such crns , the interoperability with the non - mimo crns , and many other challenging issues .
one of the main challenges in the design of crns communication protocols is the time - varying nature of the wireless channels due to the pr activities and the multi - path fading .
cooperative communication is a promising approach that can deal with the time - varying nature of the wireless channels , and hence improve the crn performance .
cooperative communication can create a virtual mimo system by allowing cr users to assist each other in data delivery ( by relaying data packets to the receiver ) .
hence , the received data packets at the cr destination traverse several independent paths achieving diversity gains .
cooperative communication can also extend the coverage area .
the benefits of employing cooperative communication , however , are achieved at the cost of an increase in power consumption , an increase in computation resources and an increase in system complexity .
it has been shown that cooperation may potentially lead to significant long - term resource savings for the whole crn .
an important challenge in this domain is how to design effective cooperative mac protocols that combine the cooperative communication with cr multiple - channel capability such that the overall network performance is improved .
the cr relay selection is another challenging problem that needs to be further investigated .
therefore , new cooperative crn mac protocols and relay selection strategies are needed to effectively utilize the available resources and maximize network performance .
network coding in crns is another interesting approach that has not yet explored in crns . based on
its verified excellent performance in wireless networks @xcite , it is natural to consider it in the design of cooperative - based crns .
the packet relaying strategies in cooperative communication are generally implemented on a per packet basis , where a store - and - forward ( sf ) technique is used ( the received packets at the cr relays are received , stored and retransmitted towards the receiver ) .
while this type of relaying mechanisms is simple , it has been shown that it provides a sub - optimal performance in terms of the overall achieved crn throughput ( especially , in multi - cast scenarios ) .
instead of using sf , network coding can be used to maximize the crn performance . with network coding , the intermediate relay cr users can combine the incoming packets using mathematical operations ( additions and subtractions over finite fields ) to generate the output packets .
one drawback in using network coding is that the computational complexity increases as the finite field size increases .
the higher the field size , the better is the network performance .
however , the tradeoff should be further investigated and more efforts are required to identify and study the benefits and drawbacks of increasing the field - size in crns .
in addition , the performance achieved through network coding can be further enhanced in crns by dynamically adapting the total number of coded packets that need to be sent by the source cr user .
such adaptation adjustment is yet to be explored , which should be based on the pr activities , link loss rates , link correlations , and nodes reachability .
cr technology has a great potential to enhance the overall spectrum efficiency . in this paper
, we first highlighted the main existing crn architectures .
then , we described the unique characteristics of their operating rf environment that need to be accounted for in designing efficient communication protocols and spectrum assignment mechanisms for these networks .
we then surveyed several spectrum sharing approaches for crns .
we showed that these approaches differ in their design objectives .
ideally , one would like to design a spectrum sharing solution that maximizes spectrum efficiency while causing no harmful interference to pr users .
we showed that interference management ( including self - coexistence and the pr coexistence ) and distributed coordination are the main crucial issues in designing efficient spectrum sharing mechanisms .
the key idea in the design of effective spectrum sharing and assignment protocols for crns is to jointly consider the pr activities and cr link - quality conditions .
the reaction to pr appearance is another important issue in designing spectrum sharing schemes for crns
. currently , most of spectrum sharing schemes are either reactive or proactive schemes .
interference threshold - based schemes are very promising , where more research should be conducted to explore their advantages and investigate their complexities .
another crucial and challenging problem is the incorporation of the guard - band constraints in the design of spectrum sharing schemes for crns .
a huge amount of interference is leaked into the adjacent channels when guard bands are not used .
this can significantly reduce spectrum efficiency and cause harmful interference to pr users .
the effect of introducing guard - bands on the spectrum sharing design has not been well explored .
many interesting open design issues still to be addressed .
variable - width spectrum sharing approach is very promising , but their design assumptions and feasibility should be carefully investigated .
resource virtualization is another important concept that can significantly improve the overall spectrum utilization .
beamforming and mimo technology have recently been proposed as a means of maximizing spectrum efficiency .
the use of beamforming in crns with mimo capability can achieve significant improvement in spectrum efficiency .
however , the spectrum sharing problem becomes more challenging due to the resurfacing of several design issues such as the determination of the beamforming weights , the joint channel assignment and power control , etc .
, which need to be further addressed .
research should focus also on the cooperative cr communication and cross - layer concepts .
using fd radios versus using hd radios is another interesting issue .
moreover , utilizing network coding is very promising in improving the crn s performance .
finally , we showed that channel bonding and aggregation can be realized through the use of d - ofdm technology .
this technology allows cr user to simultaneous transmit or receive over multiple channels using a single radio transceiver . 2 federal communications commission , spectrum policy task force , " et docket no .
02 - 135 , nov . , 2002 .
h. bany salameh and m. krunz , channel access protocols for multihop opportunistic networks : challenges and recent developments , " ieee network , vol .
23 , no . 4 , 2009 . h. bany salameh , throughput - oriented channel assignment for opportunistic spectrum access networks , " mathematical and computer modelling , vol . 53 , iss .
11 - 121 , pp .
2108 - 2118 , june 2011 .
h. salameh , rate - maximization channel assignment scheme for cognitive radio networks " , in proceedings of the ieee globecom conference , florida , 2010 . l. tan and l. le , channel assignment with access contention resolution for cognitive radio networks , " ieee transactions on vehicular technology , vol .
61 , no . 6 , pp . 2808 - 2823 , 2012 .
k . tseng , w .- h .
chung , h. chen , and c .- s .
wu , distributed energy - efficient cross - layer design for cognitive radio networks , " in proceedings of the 23rd international symposium on personal indoor and mobile radio communications ( pimrc12 ) , sydney , australia , 2012 , pp .
161 - 166 .
e. anifantis , v. karyotis , and s. papavassiliou , a markov random field framework for channel assignment in cognitive radio networks , " in proceedings of the ieee international conference on pervasive computing and communications workshops ( percom workshops12 ) , lugano , switzerland , 2012 , pp .
770 - 775 .
h. bany salameh , m. krunz , and d. manzi , an efficient guard - band - aware multi - channel spectrum sharing mechanism for dynamic access networks , " in proceedings of the ieee globecom conference , dec .
m. ahmadi , y. zhuang , and j. pan , distributed robust channel assignment for multi - radio cognitive radio networks , " in proceedings of the 76th ieee vehicular technology conference ( vtc12-fall ) , quebec city , canada , 2012 , pp . 1 - 5 . h. bany salameh , m. krunz , and d. manzi , spectrum bonding and aggregation with guard - band awareness in cognitive radio networks , " ieee transactions on mobile computing , 2014 . h. bany salameh , m. krunz , and o. younis , dynamic spectrum access protocol without power mask constraints " , in proceedings of the ieee infocom conference , brazil , april 2009 .
q. zhao , l. tong , a. swami , and y. chen , decentralized cognitive mac for opportunistic spectrum access in ad hoc networks : a pomdp framewrok , " ieee journal on selected areas in communications , vol .
589 - 600 , 2007 .
q. xiao , y. li , m. zhao , s. zhou , and j. wang , opportunistic channel selection approach under collision probability constraint in cognitive radio systems , " computer communications , vol .
18 , pp . 1914 - 1922 , 2009 .
q. zhao , s. geirhofer , l. tong , and b. m. sadler , opportunistic spectrum access via periodic channel sensing , " ieee transactions on signal processing , vol .
785 - 796 , 2008 .
h. bany salameh , m. krunz , and o. younis ,
mac protocol for opportunistic cognitive radio networks with soft guarantees , " ieee transactions on mobile computing , vol .
10 , 2009 .
wang , c .- w .
wang , and f. adachi , load - balancing spectrum decision for cognitive radio networks , " ieee journal on selected areas in communications , vol .
4 , pp . 757 - 769 , april 2011 .
p. a. k. acharya , s. singh , and h. zheng , reliable open spectrum communications through proactive spectrum access , " in proceedings of the tapas conference , 2006 .
m. bkassiny and s. k. jayaweera ,
optimal channel and power allocation for secondary users in cooperative cognitive radio networks , " in proceedings of the 2nd international conference on mobile lightweight wireless systems ( mobilight ) , 2010 .
h. wang , j. ren , and t. li , resource allocation with load balancing for cogntive radio networks , " in proceedings of the ieee globecom conference , 2010 .
u . yoon and e. ekici ,
voluntary spectrum handoff : a novel approach to spectrum management in crns , " in proceedings of the ieee icc conference , 2010 .
y. song and j. xie ,
common hopping based proactive spectrum handoff in cogntive radio ad hoc networks , " in proceedings of the globecom conference , 2010 .
o. badarneh and h. bany salameh , quality - aware routing in cognitive radio networks under dynamically varying spectrum opportunities , " computers and electrical engineering journal , vol .
38 , iss . 6 , pp . 1731 - 1744 , november 2012 .
h. bany salameh , resource management with probabilistic performance guarantees in opportunistic networks , " international journal of electronics and communications aeu , vol .
67 , iss . 7 , pp .
632 - 636 , 2013 .
h. bany salameh and o. badarneh , opportunistic medium access control for maximizing packet delivery rate in dynamic access networks , " journal of network and computer applications , vol .
523 - 532 , 2013 .
lee , s. member , and i. f. akyildiz , a spectrum decision framework for cognitive radio networks , " ieee transaction on mobile computing , vol .
161 - 174 , 2011 .
i. malanchini , m. cesana , and n. gatti , on spectrum selection games in cognitive radio networks , " in proceedings of the ieee globecom conference , 2009 , pp . 1 - 7 .
q. d. xue feng , s. guangxi , and l. yanchun , smart channel swiching in cognitive radio networks , " in proceedings of the cisp conference , 2009
. a. c .- c .
hsu , d. s .- l .
wei , and c .- c .
j. kua , a cognitive mac protocol using statistical channel allocation for wireless ad - hoc networks , " in proceedings of the ieee wcnc conference , 2007 .
hsu , and k .-
feng , a pomda - based spectrum handoff protocol for partially observable cognitive radio networks , " in proceedings of the ieee wcnc conference , 2009 .
p. zhu , j. li , and x. wang , a new channel parameter for cognitive radio , " in proceedings of the crowncom conference , 2007 .
l. yu , c. liu , and w. hu , spectrum allocation algorithm in cognitive ad - hoc networks with high energy efficiency , " in proceedings of the international conference on green circuits and systems ( icgcs ) , 2010 .
s. byun , i. balasingham , and x. liang , dynamic spectrum allocation in wireless cognitive sensor networks : improving fairness and energy efficiency , " in proceedings of the 68th ieee vehicular technology conference ( vtc 2008-fall ) , 2008 , pp . 1 - 5 .
s. gao , l. qian , and d. vaman , distributed energy efficient spectrum access in wireless cognitive radio sensor networks , " in proceedings of the ieee wcnc conference , 2008 , pp .
1442 - 1447 .
x. li , d. wang , and j. mcnair , residual energy aware channel assignment in cognitive radio sensor networks , " in proceedings of the ieee wcnc conference , 2011 , pp . 398 - 403
. t. zhang , b. wang , and z. wu , spectrum assignment in infrastructure based cognitive radio networks , " in proceedings of the ieee national aerospace and electronics conference ( naecon ) , 2009 , pp .
l. le and e. hossain , resource allocation for spectrum underlay in cognitive radio networks , " ieee transactions on wireless commuications , vol . 7 , no . 12 , pp . 5306 - 5315 , 2008 .
g. yildirim , b. canberk , and s. oktug , enhancing the performance of multiple ieee 802.11 network environment by employing a cognitive dynamic fair channel assignment , " in proceedings of the 9th ifip annual mediterranean ad hoc networking workshop ( med - hoc - net ) , 2010 , pp . 1 - 6
. y. ge , j. sun , s. shao , l. yang , and h. zhu , an improved spectrum allocation algorithm based on proportional fairness in cognitive radio networks , " in proceedings of the 12th ieee international conference on communication technology ( icct ) , 2010 , pp .
742 - 745 .
y. li , z. wang , b. cao , and w. huang , impact of spectrum allocation on connectivity of cognitive radio ad - hoc networks , " in proceedings of the ieee globecom conference , pp . 1 - 5 , dec . 2011
. m. ahmadi and j. pan , cognitive wireless mesh networks : a connectivity preserving and interference minimizing channel assignment scheme , " in proceedings of the ieee pacific rim conference on communications , computers and signal processing , pp .
458 - 463 , aug . 2011 .
h. m. almasaeid and a. e. kamal ,
receiver - based channel allocation for wireless cognitive radio mesh networks , " in proceedings of the ieee symposium on new frontiers in dynamic spectrum ( dyspan ) , pp . 1 - 10 , apr .
2010 . c. zhao , b. shen , t. cui , and k. kwak , graph - theoretic cooperative spectrum allocation in distributed cognitive networks using bipartite matching , " in proceedings of the ieee 3rd international conference on communication software and networks ( iccsn ) , 2011 , pp .
223 - 227 .
f. ye , r. yang , and y. li , genetic algorithm based spectrum assignment model in cognitive radio networks , " in proceedings of the 2nd ieee information engineering and computer science international conference ( iciecs10 ) , china , 2010 , pp . 1 - 4
. f. hou and j. huang , dynamic channel selection in cognitive radio network with channel heterogeneity , " in proceedings of the ieee globecom conference , florida , 2010 .
p. kaur , m. uddin , and a. khosla , adaptive bandwidth allocation scheme for cognitive radios , " international j. advancements in computing technology , vol .
35 - 41 , 2010 . l. gao , s. cui , power and rate control for cognitive radios : a dynamic programming approach , " in proceedings of the 3rd international conference on cognitive radio oriented wireless networks and communications ( crowncom 2008 ) , pp
. 1 - 7 , 2008 .
e. peh , y - c liang , y. guan , y. zeng ,
optimization of cooperative sensing in cognitive radio networks : a sensing - throughput trade off view " , ieee transactions on vehicular technology , vol .
5294 - 5299 , 2009 .
w. wang , j. cai , b. kasiri , and a.s .
alfa , channel assignment of cooperative spectrum sensing in multi - channel cognitive radio networks , " in proceedings of the ieee international conference on communications ( icc ) , 2011 , pp . 1 - 5 .
y. zeng , y - c liang , e. peh , a. hoang , cooperative covariance and eigenvalue based detections for robust sensing " , in proceedings of the ieee globecom conference , december 2009 , hawaii , usa .
j. unnikrishnan , v. veeravalli , cooperative sensing for primary detection in cognitive radio " , ieee journal on selected topics in signal processing , vol 2 , no .
18 - 27 , 2008 .
k. seshukumar , r. saravanan , m. suraj , spectrum sensing review in cognitive radio , " in proceedings of the ieee international conference on emerging trends in vlsi , embedded system , nano electronics and telecommunication system conference ( icevent ) , pp . 1 - 4 , jan .
s. haykin , d.j .
thomson , j.h .
reed , spectrum sensing for cognitive radio , " proceedings of the ieee , vol .
849 - 877 , 2009 .
z. quan , s. cui , a.h .
sayed , h.v .
poor , optimal multiband joint detection for spectrum sensing in cognitive radio networks , " ieee transactions on signal processing , vol .
3 , pp . 1128 - 1140 , march 2009 .
k. bian and j .- m .
park , a coexistence - aware spectrum sharing protocol for 802.22 wrans , " in proceedings of the 18th internatonal conference on computer communications and networks ( icccn ) , aug .
2009 , pp . 1 - 6
. v. gardellin , s.k .
das , and l. lenzini , a fully distributed game theoretic approach to guarantee self - coexistence among wrans , " in proceedings of the ieee infocom conference , march 2010 , pp
franklin , s .- j .
you , j .- s .
pak , m .- s .
song , and c .- j .
channel management in ieee 802.22 wran systems , " ieee communications magazine , vol .
88 - 94 , 2011 .
p. camarda , c. cormio , and c. passiatore , an exclusive self - coexistence ( esc ) resource sharing algorithm for cognitive 802.22 networks , " in proceedings of the 5th ieee international symposium on wireless pervasive computing ( iswpc ) , may 2010 , pp .
128 - 133 .
w. hu , d. willkomm , m. abusubaih , j. gross , g. vlantis , m. gerla , and a. wolisz , cognitive radios for dynamic spectrum access - dynamic frequency hopping communities for efficient ieee 802.22 operation , " ieee communications magazine , vol .
80 - 87 , may 2007 .
h. bany salameh , y. jararweh , t. aldalgamouni , and a. khreishah , traffic - driven exclusive resource sharing algorithm for mitigating selfcoexistence problem in wran systems , " in proceeding of the ieee wcnc conference , april 2014 , pp .
1933 - 1937
. s. debroy and m. chatterjee , intra - cell channel allocation scheme in ieee 802.22 networks , " in proceedings of the 7th ieee conference on consumer communications and networking conference ( ccnc ) , 2010 , pp .
284 - 289 .
m. bani hani , h. bany salameh , y. jararweh , and a. bousselham , traffic - aware self - coexistence management in ieee 802.22 wran systems , " in proceeding of the 7th ieee gcc conference and exhibition , nov 2013 , pp .
507 - 510 .
lo , a survey of common control channel design in cognitive radio networks , " physical communication , vol .
26 - 39 , 2011 .
k. chowdhury and i. akyildiz , ofdm based common control channel design for cognitive radio ad hoc networks , " ieee transactions on mobile computing , vol .
10 , pp . 228 - 238 , 2011 .
h. su and x. zhang , cross - layer based opportunistic mac protocols for qos provisionings over cognitive radio wireless networks , " ieee journal on selected areas in communications , pp .
118 - 129 , 2008 .
y. yuan , p. bahl , r. chandra , t. moscibroda , and y. wu .
, allocating dynamic time - spectrum blocks in cognitive radio networks , " in proceedings of the acm international symposium on mobile and ad - hoc networking and computing ( mobihoc ) , sept .
j. jia , q. zhang , and x. shen , hc - mac : a hardware - constrained cognitive mac for efficient spectrum management , " ieee journal on selected areas in communications , vol .
106 - 117 , jan . 2008 .
a. masri , c .- f .
chiasserini , and a. perotti , control information exchange through uwb in cognitive radio networks , " in proceedings of the @xmath8th ieee international symposium on wireless pervasive computing conference ( iswpc ) , 2010 , pp .
110 - 115 .
j. marinho , e. monteiro , corhys : hybrid signaling for opportunistic distributed cognitive radio " , computer networks , february 2015 .
park , r. chen , k. bian ,
control channel establishment in cognitive radio networks using channel hopping , " ieee journal on selected areas in communications , vol .
689 - 703 , 2011 .
u. tefek , t. lim , channel - hopping on multiple channels for full rendezvous diversity in cognitive radio networks , " in proceedings of ieee globecom conference , pp .
4714 - 4719 , dec . 2014 j .- m.j .
park , r. chen , k. bian , a quorum - based framework for establishing control channels in dynamic spectrum access networks , " in proceedings of the @xmath9th annual international conference on mobile computing and networking conference ( mobicom ) , 2009 , pp .
n. baldo , m. zorzi a. asterjadhi ,
a distributed network coded control channel for multihop cognitive radio networks , " ieee network , vol .
23 , no . 4 , 2009 , pp .
n. baldo , a. asterjadhi , and m. zorzi , dynamic spectrum access using a network coded cognitive control channel , " ieee transactions on wireless communications , vol . 9 , no . 8 , pp . 2575 - 2587 , 2010 .
c. cormio and k.r .
chowdhury ,
common control channel design for cognitive radio wireless ad hoc networks using adaptive frequency , " ad hoc networks , vol .
430 - 438 , 2010 .
dasilva and i. guerreiro , sequence - based rendezvous for dynamic spectrum access , " in proceedings of the ieee dyspan conference , 2008 , pp . 1 - 7 .
h. bany salameh , spread spectrum - based coordination design for multi - hop spectrum - agile wireless networks , " in proceedings of the ieee 81th vehicular technology conference ( vtc15-spring ) , scotland , 2015 , pp . 1 - 5
. s. perez - salgado , e. rodriguez - colina , m. pascoe - chalke , and a. prieto - guerrero ,
underlay control channel using adaptive hybrid spread spectrum techniques for dynamic spectrum access , " in proceedings of the international symposium on performance evaluation of computer and telecommunication systems , july 2013 , pp .
t. chen , h. zhang , g.m .
maggio , and i. chlamtac , cogmesh : a cluster - based cognitive radio network , " in proceedings of the ieee dyspan conference , 2007 , pp .
168 - 178 .
t. chen , h. zhang , m.d .
katz , and z. zhou , swarm intelligence based dynamic control channel assignment in cogmesh , " in proceedings of the ieee icc workshops , 2008 , pp . 123 - 128
. l. lazos , s. liu , and m. krunz ,
spectrum opportunity - based control channel assignment in cognitive radio networks , " in proceedings of the ieee secon conference , 2009 , pp . 1 - 9
. s. liu , l. lazos , and m. krunz ,
cluster - based control channel allocation in opportunistic cognitive radio networks , " ieee transactions on mobile computing , vol .
10 , pp . 1436 - 1438 , 2012 .
v. gardellin , s. das , and l. lenzini , coordination problem in cognitive wireless mesh networks , " pervasive and mobile computing , vol . 9 , no .
1 , pp . 18 - 34 , 2013 .
s. jones , n. merheb , i .- j . wang ,
an experiment for sensing - based opportunistic spectrum access in csma / ca networks " , in proceedings of the ieee dyspan conference , 2005 , pp .
593 - 596 .
j. jia , j. zhang , and q. zhang , cooperative relay for cognitive radio networks , " in proceedings of the ieee infocom conference , 2009 .
g. uyanik , m. abdel - rahman , and m. krunz , optimal guard - band - aware channel assignment with bonding and aggregation in multi - channel systems , " in proceedings of the ieee globecom conference , atlanta , usa , 2013 , pp . 4898 - 4903
. m. abdel - rahman , f. lan , and m. krunz , spectrum - efficient stochastic channel assignment for opportunistic networks , " in proceedings of the ieee globecom conference , atlanta , usa , 2013 , pp .
1272 - 1277 .
h. bany salameh and m. krunz , adaptive power - controlled mac protocols for improved throughput in hardware - constrained cognitive radio networks , " ad hoc networks journal , vol .
9 , no . 7 , pp .
1127 - 1139 , sep . 2011 .
j. r. gallego , m. canales , and j. ortin , flow allocation with joint channel and power assignment in multihop cognitive radio networks using game theory , " in proceedings of the 9th ieee international symposium on wireless communication systems ( iswcs12 ) , paris , france , 2012 , pp . 91 - 95
. j. wu , y. dai , and y. zhao , effective channel assignments in cognitive radio networks , " computer communications , vol .
411 - 420 , 2013 . c. zheng , r. p. liu , x. yang , i. b. collings , z. zhou , and e. dutkiewicz , maximum flow - segment based channel assignment and routing in cognitive radio networks , " in proceedings of the 73rd vehicular technology conference ( vtc11-spring ) , budapest , hungary , 2011 , pp . 1 - 6 y. dai and j. wu , efficient channel assignment under dynamic source routing in cognitive radio networks , " in proceedings of ieee 8th international conference on mobile adhoc and sensor systems ( mass11 ) , valencia , spain , 2011 , pp .
550 - 559 .
junior , m. fonseca , a. munaretto , a. viana , and a. ziviani , zap : a distributed channel assignment algorithm for cognitive radio networks , " eurasip journal on wireless communications and networking , vol .
2011 , no .
1 , pp . 1 - 11 , 2011 .
y. xing , c. mathur , m. haleem , r. chandramouli , k. subbalakshmi , dynamic spectrum access with qos and interference temperature constraints " , ieee transactions on mobile computing , vol .
423 - 433 , 2007 .
t. shu , s. cui , m. krunz , medium access control for multi - channel parallel transmission in cognitive radio networks , " in proceedings of globecom conference , pp . 1 - 5 , 2006 . j. vartiainen , m. hoyhtya , j. lehtomaki , and t. braysy , priority channel selection based on detection history database , " in proceedings of the crowncom conference , 2010 .
l. yang , l. cao , and h. zheng , proactive channel access in dynamic spectrum networks , " elsevier physical communications journal , vol .
103 - 111 , 2008 . t. c. clancy and b. d. walker , predictive dynamic spectrum access , " in proceedings of the sdr forum technical conference , florida , usa , 2006 .
a. mishra and r. brodersen , cooperative sensing among cognitive radios , " in proceeding of the ieee icc conference , 2006 .
x. liu and s. n. , sensing - based opportunistic channel access , " mobile networks and applications , vol .
577 - 591 , 2006 .
s. fourati , s. hamouda , and s. tabbane , rmc - mac : a reactive multi - channel mac protocol for opportunistic spectrum access , " in proceedings of the 4th ifip international conference on new technologies , mobility and security ( ntms ) , 2011 , pp . 1 - 5 .
x. wang , p. krishnamurthy , and d. tipper , wireless network virtualization " , in proceedings of the ieee international conference on computing , networking and communications ( icnc ) , pp .
818 - 822 , 2013 .
y. jararweh , m. al - ayyoub , a. doulat , a. abed al aziz , h. bany salameh , a. khreishah , software defined cognitive radio network framework : design and evaluation " , international journal of grid and high performance computing ( ijghpc ) , 2014 .
a. doulat , a. abed al aziz , m. al - ayyoub , y. jararwah , h. bany salameh , a. khreishah : software defined framework for multi - cell cognitive radio networks , " in proceedings of the ieee 10th international conference on wireless and mobile computing , networking and communications ( wimob ) , larnaca , cyprus , 2014 . q. zhao and j. ye , quickest detection in multiple on - off processes , " ieee transactions on signal processing , vol .
5994 - 6006 , dec . 2010 .
w. afifi , a. sultan , and m. nafie , adaptive sensing and transmission durations for cognitive radios , " in proceedings of the ieee dyspan11 conference , 2011 , pp .
380 - 388 .
s. huang , x. liu , and z. ding , optimal sensing - transmission structure for dynamic spectrum access , " in proceedings of the ieee infocom09 conference , april 2009 , pp .
2295 - 2303 .
w. afifi and m. krunz , adaptive transmission - reception - sensing strategy for cognitive radios with full - duplex capabilities , " on proceedings of the ieee dyspan 2014 conference , mclean , va , april 2014 . w. afifi and m. krunz , exploiting self - interference suppression for improved spectrum awareness / efficiency in cognitive radio systems . " in proceedings of the ieee infocom13 conference , turin , italy , 2013 , pp .
1258 - 1266 . w. cheng , x. zhang , and h. zhang , full duplex spectrum sensing in non - time - slotted cognitive radio networks , " in proceedings of the milcom11 conference , nov .
2011 , pp .
1029 - 1034
. m. hassan , md.j .
hossainr , cooperative beamforming for cognitive radio systems with asynchronous interference to primary user , " ieee transactions on wireless communications , vol .
11 , pp.5468 - 5479 , 2013 . c. zhang , l. guo , r. hu , j. lin ,
opportunistic distributed beamforming in cognitive radio networks with limited feedback , " in proceedings of the ieee wcnc conference , pp.893 - 897 , 2014 .
j. zhang , l. guo , t. kang , p. zhang , cooperative beamforming in cognitive radio network with two - way relay , " in proceedings of the ieee 79th vehicular technology conference ( vtc spring ) , pp . 1 - 5 ,
may 2014 . j .-
h . noh , s .-
oh , beamforming in cognitive radio with partial channel state information , " in proceedings of the ieee globecom conference , pp
. 1 - 6 , 2011 . j. poston and w. horne , discontiguous ofdm considerations for dynamic spectrum access in idle tv channels , " in proceedings of the ieee dyspan conference , 2005 , pp .
607 - 610 .
g. scutari , d. palomar , and s. barbarossa , cognitive mimo radio , " ieee signal procesing . magazine ,
46 - 59 , nov . 2008 .
l. bixio , g. oliveri , m. ottonello , m. raffetto , and c. s. regazzoni , cognitive radios with multiple antennas exploiting spatial opportunities , " ieee transaction on signal processing , vol .
58 , no . 8 , pp . 4453 - 4459 , aug .
s. hua , h. liu , m. wu , and s. panwar , exploiting mimo antennas in cooperative cognitive radio networks , " in proceeding of the ieee infocom conference , pp . 2714 - 2722 , shanghai , 2011 . | in this paper , we investigate the issue of spectrum assignment in crns and examine various opportunistic spectrum access approaches proposed in the literature . we provide insight into the efficiency of such approaches and their ability to attain their design objectives .
we discuss the factors that impact the selection of the appropriate operating channel(s ) , including the important interaction between the cognitive link - quality conditions and the time - varying nature of prns .
protocols that consider such interaction are described .
we argue that using best quality channels does not achieve the maximum possible throughput in crns ( does not provide the best spectrum utilization ) . the impact of guard bands on the design of opportunistic spectrum access protocols is also investigated .
various complementary techniques and optimization methods are underlined and discussed , including the utilization of variable - width spectrum assignment , resource virtualization , full - duplex capability , cross - layer design , beamforming and mimo technology , cooperative communication , network coding , discontinuous - ofdm technology , and software defined radios .
finally , we highlight several directions for future research in this field . |
the sfh of local group ( lg ) galaxies can be studied very accurately from cmds reaching the oldest main sequence turnoffs ( ideally complemented from spectroscopy from individual stars ) .
it provides key details on the processes of formation and evolution of these galaxies and is , therefore , a complementary path to high - redshift studies of galaxy formation and evolution .
the majority of lg galaxies are dwarfs , and dwarf galaxies are also the most numerous type in the universe .
thus , dwarf galaxies have the potential to provide important information on many key astrophysical processes . in current structure formation models ,
dwarf galaxies are the first to collapse , around 1 gyr after the big bang , and thus their oldest populations probe the conditions of the very early universe . in particular , the details of their early sfhs can provide information on the effects on galaxy formation of the reionization which occurred at high redshift .
they also can shed light on the role of other mechanisms , like supernovae feedback , in removing significant amounts of gas from small galaxy halos .
the mw satellites , the only galaxies which sfhs are known in detail , may however not be good dwarf galaxy representatives for the above purposes . because of the strong influence of our galaxy , their evolution is complicated by environmentally related mechanisms like tidal stirring ( mayer et al .
2006 ) or ram pressure stripping ( blitz & robishaw 2000 ) , which are expected to remove gas and strongly affect their sfhs .
in addition , the radiation field of the early mw is expected to dominate over the cosmological uv background ( mayer et al .
2005 ) for these reasons , the smallest * isolated * dwarf galaxies are potentially interesting cosmological probes . taking advantage of the great sensitivity and spatial resolution of acs on board the hst , we have undertaken a major project aimed at investigating the sfh of a selected sample of isolated lg dwarf galaxies .
our main goal is to obtain cmds reaching the oldest main - sequence turnoffs ( m@xmath2 in low metallicity systems ) with good photometric accuracy , for six isolated dwarf galaxies ( two dirr galaxies : leo a and ic1613 ; the two isolated dsph galaxies discovered so far in the local group : cetus and tucana ; and two transition type dirr / dsph : lgs3 and phoenix ) .
the data for this project have been obtained through two hst - acs programs ( go 10505 , p.i .
gallart ; go 10590 , p.i .
cole ) with a total of 113 awarded orbits , and will be supplemented by data from a previous hst - wfpc2 program for phoenix , ( go 8706 , p.i .
aparicio ) .
the main aims of this project are the following : @xmath3 to obtain the sfh of each galaxy through comparison of the observed cmd with synthetic cmds ( see below ) .
@xmath3 to study the radial gradients in the sfh for the galaxies in the sample , in order to gain insight on the structural evolution within dwarf galaxies .
@xmath3 to characterize the populations of variable stars in each galaxy , and to understand their relationships with the sfhs and stellar population gradients .
this will provide information on the use of variable stars as stellar population tracers . by comparing the results on isolated dwarfs with those obtained for the mw and m31 satellites , we also expect to shed new light on the effects of environmental mechanisms , which are expected to dominate for the satellite sample . a detailed comparison between cetus and tucana , and
the dsph satellites will be particularly useful to clarify to what extent these two groups of objects may be related to each other . here , we will present some preliminary results regarding the obtained cmds and the study of the sfh gradients .
other papers in this conference deal with complementary aspects , namely : the data reduction strategy ( monelli ) , the population of variable stars in tucana and lgs3 ( bernard ) , and the first quantitative derivations of the sfh in two of the galaxies : leo a ( cole ) and phoenix ( hidalgo , aparicio & martnez - delgado ) .
in this project , we have emphasized the need to obtain cmds reaching the oldest main - sequence turnoff with good photometric accuracy , in order to obtain reliable and detailed sfhs . in this case , the information can be obtained directly from the main sequence , which is the best understood phase of stellar evolution from the theoretical point of view and is also the one in which the location of stars in the cmd shows the highest sensitivity to age differences .
the range of ages and metallicities present can be determined simply through comparison with theoretical isochrones . however , to * quantitatively * determine the sfh , it is necessary to compare the observed density distribution of stars with that predicted by stellar evolution models ( see gallart , zoccali , & aparicio 2005 ) . figure [ seis ] shows the cmds of all of the galaxies in our sample .
note the variety of sfhs , as evidenced by the comparison with selected isochrones .
this is the first time that data of this high quality have been obtained for dwarf galaxies beyond the mw satellite system .
the determination of the sfhs , through comparison of observed and synthetic cmds , is underway using different approaches by different team members ( see e.g. , aparicio & gallart 2004 ; bertelli & nasi 2001 ; gallart et al .
1999 ; skillman et al .
2003 ; and references therein , for a discussion of the different approaches .
see also http://iac-star.iac.es/iac-star/ ) .
this will allow us to evaluate systematic differences in the results , and to better understand the errors affecting the sfh determinations .
see cole ( this conference , and cole et al . 2007 ) and hidalgo et al .
( this conference , and hidalgo et al . 2007 ) for the first results on leo a and phoenix respectively
. we will also pay special attention to the study of the stellar population gradients . in most cases ,
the acs covers a significant fraction of the galaxy , and therefore we will be able to determine a detailed sfh as a function of galactocentric distance .
this will provide key information on the structural evolution of each galaxy over a hubble time , shedding new light on possible formation mechanisms . in the case of ic1613 , the acs covers only a small fraction of the galaxy , and we will rely on our wfpc2 parallel field at @xmath4 12@xmath5 from the center ( _ plus _ several archival wfpc2 pointings and ground based data ; see figure [ ic1613grad]a ) to obtain information on stellar population variations .
figure [ ic1613grad]b shows the cmd of ic1613 obtained from our wfpc2 exposure . the same isochrones as in figure [ seis ] have been superimposed .
note the very little presence , if any , of populations younger than @xmath4 23 gyr .
however , a substantial amount of intermediate - age population ( as young as 3 gyr old ) exists even in this very peripheral position .
i thank my parents , juan and rosa , for coming with me and my little son daniel to yet another conference , thus allowing me to attend while sharing a special time together . funding for this project
is provided by the iac ( grant p3/94 ) , the spanish ministry of education and science ( grant aya2004 - 06343 , which includes eu feder funds ) , and nasa ( grants go-10505 and go-10590 from the stsci ) .
aparicio , a. & gallart , c. 2004 , _ aj _ , 128 , 1465 bertelli , g. & nasi , e. 2001 , _ aj _ , 121 , 1013 blitz , l. & robishaw , t. 2000 , _ apj _ , 541 , 675 cole , a.a .
et al . 2007 , _
apj _ letters , in press gallart , c. , freedman , w.l . , aparicio , a. , bertelli , g. & chiosi , c. 1999 , _ aj _ , 118 , 2245 gallart , c. , zoccali , m. & aparicio , a. 2005 , _ araa _ , 43 , 387 hidalgo , s.l . , aparicio , a. , martnez - delgado , d. 2006 , _ apj _ , submitted mayer , l. , mastropietro , c. , wadsley , j. , stadel , j. & moore , b. 2006 , _ mnras _ , 369 , 1021 mayer , l. 2005 , in iau col .
198 , near - field cosmology with dwarf elliptical galaxies , ed .
h. jerjen & b. binggeli , 220 pietrinferni , a. , cassisi , s. , salaris , m. , castelli , f. 2004 , _ apj _ , 612 , 168 skillman , e. d. , tolstoy , e. , cole , a. a. , dolphin , a. e. , saha , a. , gallagher , j. s. , dohm - palmer , r. c. & mateo , m. 2003 , _ apj _ , 596 , 253 | the star formation history ( sfh ) of local group galaxies is a powerful tool for studying their evolution , including chemical enrichment histories and stellar population gradients , which in turn may shed light on the role of reionization or supernovae feedback in galaxy formation and evolution . in particular , isolated dwarfs are ideal laboratories since their evolution has not been complicated by the vicinity of giant galaxies . in this paper
, we present the project _ local cosmology from isolated dwarfs ( lcid ) _ , aimed at deriving detailed sfhs for a sample of local group isolated dwarf galaxies . to accomplish this goal
we have collected , using the acs on board the hst , color - magnitude diagrams ( cmd ) reaching their oldest main sequence turnoff ( @xmath0 or [email protected] ) with good photometric accuracy .
some preliminary results from the cmds are shown and briefly discussed . |
although more than half of a century has elapsed since landau proposed that the charge carrier could be trapped by the distortion of a crystal lattice,@xcite and pekar introduced the term polaron,@xcite a number of questions regarding the single - polaron theory remain unanswered .
this is true even in the context of the holstein hamiltonian,@xcite one of the simplest electron - lattice coupling models for the one - electron system .
while the literature pertaining to the ground state of such a model is extensive , much less attention has been paid to the excited states , even at the low energies for which they are most interesting . the nature of the excited polaron states has been investigated within the adiabatic approximation in refs . by neglecting the polaron translation .
these works provide a simple picture of the self - trapped polaron states for strong couplings .
that is , the adiabatic softening of the phonon modes within the self - trapped polaron states results in several excitation energies below the bare phonon energy .
it follows that the lowest excitations of the system are the polaron states for which the electron and the phonons are strongly correlated .
when the polaron translation is restored , each of the soft - phonon modes below the bare phonon energy can be expected to develop the corresponding band if the local dynamics remain adiabatic .
actually , moving polarons have been described within the adiabatic approximation by neglecting the force impeding the polaron translational motion due to the lattice discreteness.@xcite however , the band structure of the spectrum was not considered in these investigations . for strong couplings ,
the bandwidth of the lowest band has been obtained by the adiabatic theory in the context of the simplest two - site model.@xcite when considering the band structure it is important to realize that the polaron states are coherent in the range of energies below the phonon threshold ( i.e. , below the minimal energy for inelastic scattering).@xcite one way to investigate their properties is by analyzing the single - electron and optical conductivity spectra . in the case of the single - polaron problem ,
the low - frequency coherent part of the spectral weight directly determines the polaron energies .
most of the previous spectral investigations,@xcite however , make no mention of the excited coherent polaron bands .
the exceptions in this respect are provided by those works employing the dynamic mean - field theory ( dmft ) , which is exact in the infinite dimension limit .
the dmft results predict one excited polaron band below the phonon threshold.@xcite for the one - dimensional system , recent exact - diagonalization ( ed ) and variational approaches@xcite obtained results for the lowest state of the first excited band .
provided that the local electron dynamics remain adiabatic , and that the adiabatic calculation of the self - trapped polaron energies gives several solutions below the inelastic phonon threshold , more than one excited polaron band is expected to occur in this energy range .
the present paper shows that the ed approaches for the infinite lattice , as implemented in recent works,@xcite can be extended to give very accurate results for the first few excited coherent polaron bands . in the range of parameters for which the method converges ,
the coherent part of the spectrum is found to exhibit up to three excited polaron bands , two more than previously described .
these results provide a better understanding of the spatial and temporal structure of the polaron states , as well as of their symmetry ( parity ) properties .
the latter can be used to rationalize the contributions of the bands to the low - energy single - electron and optical conductivity spectra , respectively .
an additional and attractive feature of the excitation spectrum obtained herein is its potential to help clarify the physical picture of the polaron crossover for moderate values of the adiabatic ratio .
it has been previously proposed,@xcite in the context of the variational analysis,@xcite that this crossover , in which the ground state evolves from a light to a heavy polaron state , can be understood as the anticrossing ( hybridization ) of two low - energy states .
this question is reconsidered here in terms of practically exact eigenstates and their aforementioned parity properties .
specifically , the hybridization is found to occur between states of equal parity , whereas states of opposite parity cross without any such hybridization .
furthermore , the hybridization of states in the crossover regime is characteristic not only of the holstein model under consideration . recently , for example , on the basis of accurate numerical results obtained by the diagrammatic quantum monte carlo method@xcite for the rashba - pekar polaron model,@xcite it has been suggested that the polaron crossover in that instance also involves a hybridization of several polaron solutions.@xcite ed methods , analogous to the one employed here , have found widespread application in the ground - state calculations of numerous other many - body problems . in this context , the present analysis of the excited states , together with their symmetry attributes , is also of interest for the description of low - energy eigenstates and spectral properties from a more general point of view .
the one - dimensional holstein model of interest here is defined by the hamiltonian @xmath0 @xmath1 is nearest - neighbor hopping energy of the electron , @xmath2 is the energy of dispersionless optical phonons , while @xmath3 is the electron - phonon coupling energy . @xmath4 and @xmath5 are creation operators for the electron and phonon at lattice site @xmath6 , respectively . as only the single - electron problem
is treated , the spin indices have been omitted . the total momentum of the system ( @xmath7 ) is the sum of electron and phonon momenta . as @xmath7 commutes with the hamiltonian ( [ ham ] ) , the complete set of states of the system can be constructed from eigenstates of @xmath8 and @xmath9 .
the hamiltonian ( [ ham ] ) is also invariant under space inversion .
however , the analysis of the resulting parity properties of the eigenstates will be deferred until later in the present discussion [ sec . [ s10d ] ] .
the eigenstates that will be considered herein are those falling in the energy window defined by @xmath10 @xmath11 is the minimal energy of the system ( the zero - momentum polaron ground - state energy ) and @xmath2 is the energy of the bare phonon excitation
. when the electron - phonon coupling is absent ( @xmath12 ) the states below the phonon threshold @xmath13 are those of the free electron band .
when @xmath3 is switched on , the free electron states evolve into those states correlated with phonons .
furthermore , additional bands can appear below @xmath13 as the coupling increases .
all states in the energy window ( [ ec ] ) are correlated and will be referred to as polaron states . for @xmath14 the electron and the additional phonons can form weakly - bound states , which results in a highly degenerate spectrum . the energies of the coherent polaron states are henceforth denoted by @xmath15 , and the wave functions by @xmath16 .
@xmath17 is the momentum of the polaron state ( which is also the system momentum ) , while @xmath18 denotes the band index .
the lowest ( ground ) polaron band will be denoted by @xmath19 . in the present paper
the term ground state is used for the state of minimal energy for a given momentum @xmath17 , and not to the @xmath20 state only .
accordingly , the first excited state corresponds to the first excited state of a given momentum @xmath17 , and so forth . for the polaron bands below the phonon threshold
@xmath13 the states @xmath16 are the eigenstates of @xmath21,@xcite @xmath22 more generally , it can be shown that for all eigenstates @xmath23 the simple sum - rule for the mean total lattice deformation , given by @xmath24 is satisfied . in eq .
( [ xtot ] ) @xmath25 is the space uncertainty of the free harmonic oscillator with frequency @xmath26 . besides its physical meaning
, the sum - rule for the mean total lattice deformation can also be used as a tool for checking the validity of results obtained with approximate polaron wave functions .
in the case of the holstein polaron problem , the ed approach uses only a finite number of states which contribute significantly to the exact polaron wave function for a given set of hamiltonian parameters . by using the hamiltonian matrix corresponding to the truncated ( reduced ) basis , and the appropriate numerical scheme ,
one calculates the polaron wave functions and energies .
the convergence of the results can be verified by increasing the number of basis states in the calculation . in most cases
the results are very accurate , provided that the truncation procedure is well chosen . the ed method developed in the current paper
is henceforth referred to as the exact translational method ( @xmath27 method ) .
the basis states of the @xmath27 method are given by@xcite @xmath28 the orthonormal wave function ( [ tstate ] ) describes an electron which is surrounded by a cloud of phonons .
the number of phonons at the @xmath29th lattice site away from the electron is given by @xmath30 , while @xmath17 corresponds to the total system momentum and @xmath31 is the lattice constant .
the basis states ( [ tstate ] ) of different momenta @xmath17 are not mixed by the holstein hamiltonian .
if the adiabatic ratio @xmath32 is not too large , the @xmath27 method can be used for studying polarons in the weak- and the strong - coupling regime , as well as in the crossover regime between them .
the present paper is focused , particularly , on the regime in which @xmath33 and @xmath3 is arbitrary . for @xmath34 ,
the number of relevant basis states ( [ tstate ] ) becomes large and the problem of finding the polaron wave functions becomes untreatable .
unfortunately , the limit @xmath34 which is not considered here , is difficult for other known numerical methods as well .
the matrix representation of the holstein hamiltonian in the @xmath27 basis leads to a sparse matrix . that is , it is straightforward to show that , by acting on the state ( [ tstate ] ) with the hamiltonian ( [ ham ] ) , the maximum number of non - zero matrix elements per @xmath27 basis state is five .
as an example , let us form the reduced hilbert space of only five @xmath27 states , i.e. , of the zero - phonon state @xmath35 , of three states with one phonon ( phonon at the electron site , at the left site from the electron , and at the right site from the electron ) , @xmath36 and finally , of the state with two phonons at the electron site @xmath37 .
the corresponding hamiltonian matrix is given by @xmath38\;.\ ] ] in the limit @xmath39 the @xmath27 method gives a good polaron ground state with only these five states .
nevertheless , for @xmath40 one usually has to work with a truncated basis of quite large dimension .
in other words , the @xmath27 method generally requires a numerical scheme capable of dealing with large sparse matrices . for this purpose
the well - known lanczos algorithm appears to be the most appropriate choice.@xcite indeed , the previous papers using the @xmath27 method for ground - state calculations have employed this technique .
an additional , attractive , feature of the lanczos algorithm is that it is capable of finding not just one , but rather a few extreme eigenvalues ( and eigenvectors ) of sparse matrices , provided they are well separated ( lying in the discrete part of the spectrum ) .
accordingly , the current results have been obtained by the lanczos procedure with the so - called local orthogonalization .
in addition , the states have been calculated by the block - lanczos procedure , with both variants giving the same results .
it should be stressed that the present results are compared to the already cited ground - state results@xcite of the @xmath27 method .
all states are checked through the sum - rule ( [ xtot ] ) , as well as through their mutual orthogonality .
at the beginning of our discussion it is instructive to observe how the polaron bands are formed in regard to the strength of the electron - phonon coupling @xmath3 . for this purpose ,
the polaron bands are plotted in the four panels of fig.[fig1 ] as functions of the momentum ( @xmath17 ) , for @xmath41 and different values of @xmath3 . as in the remainder of the paper
, @xmath42 is used as the energy unit . notice that , as discussed in connection with eq .
( [ ec ] ) , only the part of the spectrum below the phonon threshold ( between @xmath43 and @xmath13 ) is shown . in the weak - coupling limit for the lowest band ( @xmath19 ) ,
two regimes exist with respect to the critical momentum @xmath44.@xcite for @xmath45 the polaron state of energy @xmath46 is the ground state of the system .
for @xmath47 the ground state consists of the polaron and the unbound phonon excitation .
the unbound phonon excitation carries the system momentum @xmath17 , while the polaron momentum is equal to @xmath48 . for @xmath47
the ground - state energy lies at the bottom of the incoherent part of the spectrum @xmath14 .
the first panel of fig .
[ fig1 ] shows the @xmath27 weak - coupling results .
notice that the top flat part of the lowest band for @xmath47 has been cut by the frame .
the reason is that the numerical error of the corresponding @xmath27 states is slightly greater than that of the zero - momentum ground state , while the energy interval shown is exactly equal to the bare phonon energy @xmath2 .
as the electron - phonon coupling increases , the lowest band becomes more renormalized , and finally , at some critical coupling , the whole band falls bellow @xmath13 . by increasing electron - phonon coupling further ,
the additional polaron bands emerge below @xmath13 . in the second and the third panel of fig .
[ fig1 ] the results for @xmath49 and @xmath50 are shown , respectively .
these are the choices of parameters that correspond to the crossover regime . for @xmath49 ,
only a part of the first excited band ( @xmath51 ) lies below @xmath13 , whereas for @xmath50 there are four polaron bands in the relevant energy window . from the third panel of fig .
[ fig1 ] ( @xmath50 ) one sees that the top of the second excited band ( @xmath52 ) is at @xmath20 , and the bottom is at @xmath53 .
in addition , this band ( @xmath52 ) crosses the other excited bands ( @xmath54 ) , i.e. , the first excited band near the end of the brillouin zone ( for @xmath55 ) , and the third excited band near the center of the brillouin zone . for strong couplings ( @xmath56 )
the @xmath27 results are shown in the last panel of fig .
all the bands are very narrow .
although it is hard to distinguish the third excited band ( @xmath57 ) from the plot - frame , note that @xmath58 for all @xmath17 .
figure [ fig2 ] gives further insight into the polaron band formation .
here , the @xmath20 and @xmath53 energy curves , shifted by @xmath43 , are plotted for four bands ( @xmath18 is the band index ) .
the results are presented for the crossover and the strong - coupling regimes . in the weak - coupling regime (
not shown on fig .
[ fig2 ] ) @xmath59 becomes smaller than @xmath13 for @xmath60 .
it can be seen from fig .
[ fig2 ] that the bandwidth of the lowest band decreases continuously with @xmath3 .
on the other hand , the bandwidth of the first excited state shows a more complicated behavior .
namely , for @xmath61 this bandwidth is maximal , while , unlike for the lowest band , it decreases for smaller values of @xmath3 .
its maximal value and the minimum of @xmath62 ( @xmath63 ) correspond to similar electron - phonon couplings . from fig .
[ fig2 ] the band crossing which involves the second and the two other excited bands can be clearly seen to occur in the crossover regime .
that fact that these bands cross each other indicates that they belong to different symmetries , as will be discussed further below [ sec .
[ s10d ] ] . in strong - coupling regime
the spectrum consists of very narrow bands .
although in fig .
[ fig2 ] the bands are well separated , for very large couplings all of the excited bands approach @xmath13 from below .
finally , it can be seen from fig .
[ fig2 ] that , for @xmath41 , the energy of the third excited band stays close to @xmath13 both in the crossover and the strong - coupling regimes .
the @xmath17 values corresponding to the top of the third excited band depend sensitively on the hamiltonian parameters .
it should therefore be kept in mind that , in contrast to the other bands , the @xmath53 curve only approximately determines the boundary of this band . in particular , the third excited band is slightly wider than suggested by fig .
[ fig2 ] in the crossover regime . in order to understand the physical background of the bands shown in figs .
[ fig1 ] and [ fig2 ] , the strong - coupling and the crossover regime are examined separately in the following sections . in the strong coupling regime
the time scale relevant to the polaron translation is much slower than the time scale ( @xmath64 ) involved in the local polaron dynamics .
therefore , the local interplay between the electron density and the lattice deformation is almost independent of the polaron momentum . from the energy point of view
the contributions corresponding to the polaron translation can be neglected , and the polarons can be treated , to a good approximation , as self - trapped . the deep potential well of the lattice deformation of the self - trapped polaron captures the electron . when the electron is light ( @xmath65 ) , it is able to follow the slow motion of the lattice deformation ( e.g. , zero - point motion ) , which results in the adiabatic renormalization of the phonon modes within the lattice deformation .
the physical picture of the aforementioned self - trapped polaron follows from the adiabatic theory , and can be considered to be well understood .
thus , in order to achieve a better understanding of the current numerical results , it is convenient to compare the renormalized phonon energies obtained in the adiabatic limit to the spectrum calculated by the @xmath27 method . for the holsteinpolaron problem
the renormalized phonon modes have been calculated by different adiabatic approximations.@xcite the procedure of ref .
( born - oppenheimer approximation therein sec .
iii ) treats the lattice discreteness directly , while the polaron translation is neglected .
this approximation is appropriate for the present case as the lattice discreteness is important for the small self - trapped polarons , while the polaron translation has only a minor contribution to the energy .
the three short - dashed curves in fig .
[ fig3 ] are the renormalized phonon energies of ref . .
the lowest excitation is a symmetric vibration of the lattice with respect to the polaron center ( breathing mode ) .
the next excitation is the antisymmetric vibration ( pinning mode ) .
the third excited mode is again a symmetric vibration , although extended over a larger number of lattice sites than in the case of the lowest excitation . in fig.[fig3 ] , the solid curves are the @xmath27 energies of the @xmath20 polaron excited states . in the strong coupling limit [ the right part of fig .
[ fig3 ] ] the results of ref .
are recovered asymptotically , from below . the inset of fig
. [ fig3 ] , where only a small part of the spectrum below @xmath13 is shown , clearly demonstrates this behavior . for strong couplings ,
the positions of the excited bands are given by the the symmetric and antisymmetric phonon excitations of the adiabatic theory .
in other words , the local dynamics of the self - trapped polarons is adiabatic .
the energy @xmath66 is plotted as the long - dashed curve in fig.[fig3 ] .
this curve defines the bandwidth of the lowest polaron band which can be seen to become large in the crossover regime [ the left part of fig.[fig3 ] ] , as the translation of the polaron becomes important . on the other hand , in the strong coupling limit
, the @xmath27 method reproduces the narrow cosine polaron bands , @xmath67\;,\ ] ] where @xmath68 is an effective polaron nearest - neighbor hopping energy for the @xmath18th excited state . for strong couplings , the values of bandwidths may differ considerably from band to band . in fig .
[ fig4 ] the bandwidths , as functions of @xmath3 , are compared for constant @xmath69 .
the solid curves are the results for the three lowest polaron bands .
one sees that the effective hopping @xmath68 is increased if the polaron ( i.e. , the local lattice deformation ) is excited .
however , for strong couplings all bands are very narrow ( @xmath70 ) , meaning that the polarons are very heavy ( self - trapped ) . in real solids the coherent transport of such band states can be destroyed by the polaron - polaron interaction or by imperfections.@xcite a close inspection of the @xmath27 results in fig .
[ fig4 ] reveals that the hopping energy @xmath71 of the self - trapped polaron for the lowest band is given by the simple relation @xmath72}\;.\label{tpol}\ ] ] equation ( [ tpol ] ) is an extension of the well - known expression for the nonadiabatic small - polaron hopping energy , which is obtained by setting @xmath73.@xcite @xmath74 turns out to be a function of the adiabatic parameter @xmath75 only , i.e. , @xmath74 is an adiabatic correction to the nonadiabatic small - polaron hopping .
notice that in the adiabatic limit @xmath75 defines the spreading of the lattice deformation ( the polaron size @xmath76 ) , as well as the renormalized phonon energies.@xcite for @xmath77 the excellent fit to the @xmath27 results in fig .
[ fig4 ] ( the short - dashed curve that follows almost exactly the @xmath27 results for the the lowest band ) is achieved with @xmath78 , even though the bandwidth of the lowest polaron band changes by several orders of magnitude . in summary , for the parameters under consideration [ fig .
[ fig4 ] ] , it follows from eq .
( [ tpol ] ) that the translational dynamics of the lowest band are essentially nonadiabatic , with adiabatic corrections .
the narrow bandwidth defines the slow time scale for the polaron hopping .
in contrast , fig .
[ fig3 ] shows that the fast local dynamics of the self - trapped polarons is adiabatic .
the adiabatic softening of the local phonon modes determines the positions of the ground and excited bands in the spectrum .
the energy shift due to the softening is much larger than the bandwidths , which means that the polaron hopping can be neglected for the local dynamics . in spite of the fact that the almost exact results for the lowest polaron band in the crossover regime , obtained by various numerical methods , have been known for quite some time
, the corresponding qualitative explanation of the polaron properties does not appear completely satisfactory .
it has been suggested in ref .
, by the use of global - local method,@xcite that a simple empirical relation between hamiltonian parameters @xmath79 describes the regime for which the fast change from light to heavy polaron ground state takes place .
indeed , @xmath80 predicts , to a good approximation , the set of parameters for which the variation of the polaron effective mass ( as a function of @xmath3 ) is the fastest .
for @xmath81 the crossover occurs in the adiabatic regime when @xmath82 , i.e. , @xmath83 .
approaching @xmath84 from the strong - coupling adiabatic side ( @xmath85 ) , the size of the polaron increases .
consequently , the peierls - nabarro ( pn ) barrier decreases and the tunneling of the adiabatic polaron to the neighboring sites becomes possible .
this effect is also responsible for the coupling of the pinning and the breathing lattice modes.@xcite at @xmath86 , the restoring force of the pinning mode due to the pn barrier can be neglected , and one may treat the polaron as a freely moving.@xcite such a scenario thus describes the crossover from the self - trapped ( heavy ) to propagating ( light ) polaron states in the adiabatic limit .
on the other hand , for @xmath87 , the nonadiabatic theory describes the polaron translation for arbitrary @xmath3 , and in particular , at @xmath88 of eq.([gst ] ) . in this case
, the polaron crossover appears essentially as a passage from the weak- to the strong - coupling ground state . a reinspection of fig .
[ fig2 ] shows that the two well - separated time scales found in the strong - coupling regime , one related to the polaron translation ( given by the bandwidth ) , and the other related to the polaron local dynamics ( given by the renormalized phonon energies ) , become comparable in the crossover regime .
furthermore , the behavior of the ground and excited bands for @xmath89 indicates that the polaron translation plays an important role in the excitation spectrum . in other words ,
the influence of the polaron hopping to the neighboring sites on the renormalized modes can not be neglected under these conditions , as it can for strong couplings .
the substantial difference in the crossover regime [ see fig . [ fig3 ] ] between the @xmath27 results and the ( adiabatically renormalized ) phonon energies of ref .
is explained in this way .
this difference , however , is not necessarily related to nonadiabatic effects , as for the moderate values of the adiabatic ratio under current consideration ( @xmath90 ) it is not clear to what extent the nonadiabatic dynamics enter into the description of the polaron translation . irrespective of their temporal ( or spatial ) properties , when the polaron bands cross and/or anticross as in fig .
[ fig2 ] , their symmetry properties become important . as already mentioned
, the holstein hamiltonian is invariant under the space inversion , and its eigenstates can be distinguished according their parity .
a simple linear transformation relates the momentum @xmath16 and parity @xmath91 eigenstates , @xmath92 so that @xmath93 @xmath94 denotes the space inversion operator , and @xmath95 are the even- and odd - parity eigenvalues , respectively .
considering the @xmath27 method , the parity can be directly determined by inspection of the wave function properties .
the eigenstates @xmath16 can be expanded in terms of the basis states ( [ tstate ] ) with expansion coefficients @xmath96 , latexmath:[\[|\psi_k^{(i)}\rangle=\sum_{n_i}a_{n_0 , n_{-1 } , n_1, ... ,n_m}\ ; @xmath94 acting on the basis state ( [ tstate ] ) gives @xmath98 consequently , using eq .
( [ parity0 ] ) , and the fact that @xmath99 is complex conjugate of @xmath16 ( time - reversal ) , one finds that the expansion coefficients satisfy @xmath100 @xmath101 denotes the complex conjugate of @xmath31 .
the expansion coefficients in eq .
( [ parity ] ) stand for two local phonon configurations , the first one is obtained from the second one when the phonons to the left and right of the electron are interchanged ( @xmath102 ) . applying the above analysis within the framework of the @xmath27 method yields @xmath103 for the second excited band ( @xmath52 ) , while @xmath104 for the ground and the other two lowest excited bands ( @xmath105 ) . in this respect ,
the polaron bands inherit the symmetry of the renormalized phonon modes as obtained in the adiabatic limit . in sec .
[ s10b ] it was pointed out that the pinning mode corresponding to the second excited band is an antisymmetric vibration of the lattice ( having odd parity ) , whereas the other two lowest excited modes are symmetric vibrations ( having even parity ) .
furthermore , one sees that the crossing ( rather than the anticrossing ) between the excited bands in the crossover regime , shown in fig .
[ fig2 ] , involves bands of opposite symmetry under the space inversion . for @xmath20 and @xmath53 the linear transformation@xcite of the basis ( [ tstate ] )
@xmath106 defines two subspaces with different parities .
it follows from eq.([parity ] ) that the @xmath104 eigenstates @xmath107 belong to the @xmath108 subspace of eq .
( [ pmstate ] ) , while the @xmath103 eigenstates belong to the @xmath109 subspace .
one sees that for @xmath20 and @xmath53 the parity actually defines the symmetry of the local phonon configuration with respect to the electron . for @xmath104
this configuration is symmetric , while it is antisymmetric for @xmath103 . in ref .
the correlated behavior of the ground and the first excited state in the crossover regime was analyzed for moderate values of the adiabatic ratio using a method based on the variational approach .
it was argued that the anticrossing of two physically different polaron states , one ( heavy ) for which the translation energy is almost negligible , and the other ( light ) for which this energy is important , can describe in simple terms the mechanism of the crossover .
it follows from this interpretation that by increasing @xmath3 in the critical region of parameters near @xmath80 [ eq.([gst ] ) ] , the contribution to the ground state from the light state decreases in favor of the heavy state which has lower energy at stronger couplings . on the other hand ,
the opposite change occurs for the first excited state , which is heavier than the ground state for @xmath110 , while being lighter for @xmath111 . in fig .
[ fig5 ] the polaron ground and first excited energies obtained in ref . , denoted by @xmath112 ,
are compared to those of the @xmath27 method .
although the @xmath112 energies may be considered as a fair approximation , the anticrossing picture can be discussed more accurately in the context of the current @xmath27 results .
one sees that the @xmath27 results confirm that the minimum of @xmath62 , the two states of the same symmetry ( @xmath104 ) , corresponds to @xmath89 .
it is important to notice that the @xmath103 states of the second excited band ( @xmath52 ) [ fig .
[ fig2 ] ] are not involved in the anticrossing .
the rest of the investigation presented here is focussed on polaron effective mass and on the properties of the lattice deformation .
the polaron effective mass @xmath113 can be evaluated numerically by using the relation@xcite @xmath114 where @xmath115 is a small finite deviation of the momentum from the value @xmath20 .
@xmath116 is the effective mass of the electron , and @xmath18 denotes the band number .
the ground- ( @xmath117 ) and the first excited - state ( @xmath118 ) effective masses are compared for @xmath119 in fig .
although @xmath120 for @xmath121 , @xmath118 becomes smaller than @xmath117 as @xmath3 decreases . for larger @xmath1 ,
the results are qualitatively the same [ e.g. , see fig . [ fig9 ] ] , which means that the anticrossing picture indeed matches the behavior of @xmath117 and @xmath118 near @xmath80 . for fixed @xmath3 ,
the mean total lattice deformation ( [ xtot ] ) is the same for the ground and for the excited states .
however , the local phonon cloud around the electron can be more or less localized , which affects the polaron hopping to the neighboring sites .
a light state implies that the associated local lattice deformation is spread to a larger number of lattice sites than for the heavy state .
namely , such a deformation gives rise to a greater effective hopping integral .
the mean number of phonons in the polaron state is given by @xmath122 in the adiabatic limit , this number is quadratic in the local lattice deformation for the ground state .
a more localized ( heavier ) lattice deformation leads to larger @xmath123 . within the adiabatic approximation
the excited renormalized phonon should , in general , increase the mean number of phonons with respect to @xmath123 .
particularly , in the strong - coupling limit one obtains @xmath124 as the renormalization of the lattice vibrations becomes negligible . from fig .
[ fig6 ] , in which @xmath125 is plotted as a function of @xmath3 , one sees that the @xmath27 results tend to eq.([np1 ] ) for strong couplings . in the crossover regime ,
on the other hand , @xmath125 deviates from eq .
( [ np1 ] ) considerably .
@xmath126 is greater than @xmath127 for @xmath110 and smaller than @xmath127 for @xmath111 . for larger @xmath1 ,
the amplitude of the deviation increases even further . as shown in in the inset of fig.[fig7 ] , @xmath128 for @xmath129 , while the minimal and maximal values of @xmath125 define an interval of almost 5 phonons .
additional insights into the local polaron properties can be provided by studying the appropriate electron - lattice correlation function , @xmath130 @xmath131 is the mean lattice deformation induced at the @xmath6th site away from the electron .
the correlation function ( [ xn ] ) is normalized in such a way that @xmath132 . as the correlation spreads to adjacent lattice sites , the correlation at the electron site @xmath133 decreases .
consequently , @xmath133 should be larger for the heavy than for the light polaron state .
the ground- and first excited - state results for @xmath134 are plotted in fig .
the relationship between @xmath135 and @xmath136 changes near @xmath80 .
this is exactly what one would expect from the anticrossing picture .
@xmath137 for @xmath110 . as
the contribution of the heavy state to the ground state becomes dominant for @xmath111 , the relationship changes , and @xmath138 .
the spreading of the lattice deformation as a function of @xmath3 may also be deduced from @xmath139 and @xmath140 .
the results corresponding to this lattice site , shown in fig .
[ fig7 ] , again lead one to the same conclusion . for @xmath110 the lattice deformation seems to spread more for the ground than for the excited state , and vice - versa for @xmath111 .
all of the aforementioned ground- and first excited - state properties ( the energy , the effective mass , the mean number of phonons , and the electron - lattice deformation correlation function ) indicate that for @xmath141 the anticrossing of two ( light and heavy ) polaron states occurs near @xmath80 .
although criteria such as @xmath142 being minimal , @xmath143 , @xmath144 , or alternative criteria , do not agree exactly , they all predict the polaron crossover to occur within the same very narrow parameter range , given approximately by eq .
( [ gst ] ) .
denoting the vacuum state by @xmath145 , the zero - temperature single - electron green function can be expressed as@xcite @xmath146 for the energies below the phonon threshold @xmath147 no polaron - phonon scattering takes place , which has the consequence that the imaginary part of the electron self - energy @xmath148 tends to zero.@xcite accordingly , the low - energy part of the spectral function @xmath149 , denoted by @xmath150 , is defined by the simple poles of @xmath151 at @xmath152 , @xmath153 i.e. , the spectral function @xmath150 is defined by the polaron energy @xmath154 and the quasiparticle weight @xmath155 . in particular , for @xmath20 and @xmath53 , the quasiparticle weight @xmath156 vanishes by parity for the @xmath103 eigenstates @xmath107 , because the free electron ( zero - phonon ) state @xmath157 belongs to the @xmath108 subspace of eq .
( [ pmstate ] ) . in fig .
[ fig8 ] @xmath158 is plotted as a function of @xmath159 for a few different values of @xmath3 in the crossover regime .
the spectral weight corresponding to the ground polaron state is of the same order of magnitude as the one corresponding to the excited ( even - parity ) polaron states .
moreover , @xmath160 defined by @xmath161 reveals that , in the crossover regime ( for @xmath110 ) , nearly 40 percent of the total spectral weight confined to the ground and excited polaron states can be located in the energy window below the phonon threshold .
therefore , at least for @xmath162 and for the values of adiabatic ratio under current considerations ( @xmath141 ) , the results imply the existence of a few well - pronounced peaks in the single - electron spectral density below the phonon threshold . for stronger couplings , on the other hand ,
the quasiparticle weight below the phonon threshold is almost completely suppressed . when the electron propagator @xmath151 is expressed in terms of the electron self - energy @xmath163 , both @xmath164 and @xmath154 of eq .
( [ ak ] ) can be related to @xmath163 .
e.g. , the polaron effective mass @xmath113 is given in terms of the self - energy @xmath163 by the standard formula@xcite
@xmath165 where @xmath166 , and @xmath167 . in eq .
( [ meff2 ] ) @xmath18 is used to distinguish between results for the ground and excited ( even - parity ) states .
@xmath168 is the abbreviation for the appropriately normalized @xmath17-dependent contribution to @xmath113 .
when the self - energy @xmath163 is local ( @xmath17-independent , @xmath169 ) the polaron effective mass is related solely to the quasiparticle weight @xmath170 .
thus , the non - local character of the self - energy can be investigated by comparing @xmath168 to unity.@xcite such analysis is particularly interesting in the context of dmft , as the locality ( implying @xmath169 ) of the self - energy is an essential ingredient of this approach.@xcite in both , the weak - coupling ( @xmath171 ) and nonadiabatic ( @xmath172 ) limits , @xmath163 becomes local.@xcite within the holstein - lang - firsov approximation@xcite and for @xmath172 , the self - energy @xmath173 is , to a good approximation , given by its expansion to the first order in @xmath159 over the whole energy range defined by the lowest band . in this case
, @xmath174 defines the renormalization of the narrow cosine - like lowest band for any @xmath17 . with the @xmath27 method @xmath168
can be obtained in terms of @xmath170 and @xmath113 , the latter being estimated numerically from the band dispersion at the center of the brillouin zone [ eq .
( [ meff ] ) ] .
the @xmath27 results agree with the aforementioned analytical findings for small @xmath1 and/or @xmath3.@xcite it follows that significant non - local contributions to @xmath163 can possibly occur in the regime where neither @xmath1 nor @xmath3 are small . in the left panel of fig .
[ fig9 ] @xmath170 is plotted for moderate @xmath1 as function of @xmath3 , while the right panel shows @xmath168 for the same set of parameters .
although @xmath168 does not contribute to @xmath113 substantially for @xmath110 , at stronger couplings @xmath168 reaches values of the order of unity for both the ground and the first excited state .
that is , the non - local contribution to the electron self - energy is as equally important as the local one .
however , it should be noticed that , in this regime , the quasiparticle weight @xmath175 becomes small .
the interband optical conductivity may be evaluated within the linear - response theory from the current - current correlation function .
the low - energy ( coherent ) contribution to the real part of the interband conductivity at zero temperature is given by@xcite @xmath176 where @xmath177 are the @xmath20 polaron states , herein calculated by the @xmath27 method .
the incoherent contribution to the interband conductivity , denoted by @xmath178 , corresponds to the continuum of the excitation spectrum above the phonon threshold . since the current operator in eq .
( [ sk ] ) , defined by @xmath179 is odd under the space inversion , the nonvanishing matrix elements in eq .
( [ sk ] ) are those between the @xmath104 ground state @xmath180 and @xmath103 excited states @xmath177 .
actually , the incoherent part of the interband optical conductivity @xmath178 involves only @xmath103 excitations as well . on the contrary , as argued in eq .
( [ ak ] ) , the spectral function @xmath149 contains information about the @xmath104 part of the spectrum at @xmath181 .
the optical conductivity at zero temperature is therefore described by those excited states for which the quasiparticle weight @xmath182 vanishes , i.e. , those states that are not seen in the single - electron spectral function .
the real part of the optical conductivity @xmath183 includes the interband part and the ( intraband ) drude term at @xmath184 .
it follows from the well - known sum - rule@xcite that the total spectral weight of @xmath183 is given by the mean value of the electron kinetic energy in the ground state .
consequently , not only is the @xmath27 method capable of calculating @xmath185 , but the total spectral weight of @xmath183 is also accessible .
furthermore , one may estimate the weight of the drude term from the polaron effective mass.@xcite according to the @xmath27 results , for @xmath186 there is only one interband transition @xmath187 which contributes at zero temperature to @xmath185 [ eq . ( [ sk ] ) ] .
it is interesting to find the corresponding non - drude spectral weight below the phonon threshold .
however , @xmath188 is less than @xmath13 for relatively strong couplings , for which most of the spectral weight of @xmath183 belongs to high energies . in this parameter regime
only a few percent ( or less ) of the total spectral weight of @xmath183 corresponds to @xmath185 , or to the drude term .
this simple low - energy picture of @xmath183 becomes certainly more complex at finite temperature , particularly in the crossover regime where the bandwidths of the coherent polaron states increase .
studies at finite temperature , however , require different methods than the one presented here .
this paper reports an exact - diagonalization study of the ground and excited polaron states in the one - dimensional holstein model . for values of the adiabatic ratio @xmath141 ,
accurate energies and wave functions are obtained for translationally invariant solutions of the infinite lattice problem .
the chosen method is restricted to the part of the spectrum below the phonon threshold , for which there are no phonon excitations uncorrelated to the polaron .
the eigenstates of the holstein hamiltonian can be distinguished according to their parity .
this property follows form the hamiltonian which is invariant under the space inversion .
it is shown that only the odd excited states of the system are relevant to the optical conductivity at zero temperature . however , the contribution of the coherent excited polaron state is rather small with respect to the total spectral weight . on the other hand ,
it is the even states that contribute to the single - electron spectral function for @xmath181 . in this case
, the contribution of the coherent excited polaron states is found to be important , particularly in the crossover regime . in the strong - coupling regime
the results agree with the picture of self - trapped polarons , which may be regarded as well understood .
the spectrum is characterized by very small polaron bandwidths .
consequently , the time scale associated with the polaron translation is very large , and the contributions to the fast local polaron dynamics from the polaron hopping are negligible . by making a comparison with the @xmath27 results
, it is shown that the adiabatic theory ( the born - oppenheimer approximation ) provides a good description of the local properties of the self - trapped polaron .
namely , the energies of the renormalized phonon excitations are close to the @xmath27 excited - band energies
. two of the excited bands correspond to the symmetric adiabatic vibrations of the lattice , while one corresponds to the antisymmetric ( pinning ) vibration .
finally , it is found that for strong couplings the electron self - energy shows significant non - local ( @xmath17-dependent ) behavior at moderate @xmath1 .
as the electron - phonon coupling decreases at moderate @xmath1 , the simple band structure found in the strong - coupling regime ( i.e. , the narrow and well separated polaron bands ) evolves notably .
the effective hopping integral of the ground and excited polaron states defines a time scale which , in the crossover regime , is comparable to the time scale relevant for the local polaron dynamics . in the crossover regime
the excited ( odd - parity ) band which corresponds to the pinning vibration in the strong - coupling regime , crosses the other two ( even - parity ) bands which correspond to symmetric vibrations .
the results suggest that the same critical set of parameters defined by @xmath80 [ eq . ( [ gst ] ) ] , found for the polaron ground - state crossover , may also be associated with the rapid change of the low - energy excitation spectrum . aside from the fact that the polaron translation and the local dynamics mix in the crossover regime , the intermediate ( moderate ) values of the adiabatic ratio employed herein present additional difficulties for qualitative understanding .
namely , for @xmath33 , both the adiabatic and the nonadiabatic contributions could be important for the polaron crossover . nevertheless , in the context of the @xmath27 approach , the analyses of the energy , the effective mass , and the lattice deformation properties , all agree in one important respect . that is , for moderate values of the adiabatic ratio , two polaron states of even parity ( one heavy and one light ) undergo an anticrossing in the same region of parameters for which the ground state crosses over from the weak- to the strong - coupling regime .
this work was supported in part by the project scopes-7kpj065619 of the swiss national science foundation .
the calculations were performed mainly on the pc cluster system of the institute of physics in zagreb , donated by the alexander von humboldt foundation . | an exact diagonalization technique is used to investigate the low - lying excited polaron states in the holstein model for the infinite one - dimensional lattice . for moderate values of the adiabatic ratio , a new and comprehensive picture , involving three excited ( coherent ) polaron bands below the phonon threshold ,
is obtained .
the coherent contribution of the excited states to both the single - electron spectral density and the optical conductivity is evaluated and , due to the invariance of the hamiltonian under the space inversion , the two are shown to contain complementary information about the single - electron system at zero temperature .
the chosen method reveals the connection between the excited bands and the renormalized local phonon excitations of the adiabatic theory , as well as the regime of parameters for which the electron self - energy has notable non - local contributions .
finally , it is shown that the hybridization of two polaron states allows a simple description of the ground and first excited state in the crossover regime . |
the conversion of molecular gas into stars and planets can for simplicity be conceptually divided into four stages : ( 1 ) the fragmentation of clouds into large scale structures ( filaments / ribbons ) ; ( 2 ) the formation of dense cores within these large structures ; ( 3 ) core collapse to form a hydrostatic protostar ; ( 4 ) disk formation and evolution leading to outflows , multiplicity , planets , etc . in this review
i emphasize step ( 1 ) , given recent developments in the literature , and comment briefly on the others .
molecular clouds are assembled from h i clouds in the interstellar medium that have a mass - to - flux ratio @xmath0 that is significantly less than the critical value required for gravitational fragmentation and collapse ( ( * ? ? ?
* heiles & troland 2005 ) ) .
if molecular clouds are assembled from these subcritical components , then the observed star formation in the clouds leads to two possibilities : ( 1 ) the molecular clouds are entirely supercritical , having been assembled from a sufficient amount of subcritical material , by motions along galactic magnetic field lines ; ( 2 ) the molecular clouds are largely subcritical but that the breakdown of flux freezing due to ambipolar diffusion allows subregions to become subcritical and produce dense cores . the first scenario places challenging demands on the accumulation length of matter in molecular clouds , and also on the speed of streaming motions along the magnetic field ( ( * ?
* mestel 1999 ) ) . the second scenario has has the drawback that the ambipolar diffusion timescale is at least @xmath1 yr , or even much longer , in diffuse molecular gas .
however , the addition of turbulent initial conditions leads to the formation of dense structures on a much shorter turbulent crossing time . within the compressed gas
, rapid ambipolar diffusion can take place due to the elevated density and magnetic field gradient .
this scenario has been explored in many studies over the last decade ( ( * ? ? ?
* nakamura & li 2005 ) ; ( * ? ? ?
* elmegreen 2007 ) ; ( * ? ? ?
* kudoh & basu 2008 ) ; ( * ? ? ?
* nakamura & li 2008 ) ; ( * ? ? ?
* basu et al . 2009 ) ; ( * ? ? ?
* kudoh & basu 2011 ) ) and also provides a natural way to understand filamentary structures in clouds . by filaments
we do not mean objects with cylindrical symmetry , but rather more ribbon - like structures , with flattening along the ambient magnetic field direction , and compression by turbulence in primarily one direction perpendicular to the magnetic field .
the overall scenario is illustrated in figure [ scenario ] , taken from a simulation of ( * ? ? ?
* basu et al . ( 2009 ) ) .
the filament formation is thus quite distinct from the cosmological filament formation simulations ( e.g. , ( * ? ? ?
* springel et al . 2005 ) ) .
in which the gravitational field of cold dark matter is able to collapse into filamentary structures on a local dynamical time .
the magnetic turbulent picture allows the formation of filaments but not necessarily in a volume filling network as in cosmological simulations , since large parts of clouds may remain in a subcritical common envelope with little to no star formation .
models of magnetohydrodynamic ( mhd ) wave propagation in inhomogeneous clouds ( ( * ? ? ?
* kudoh & basu 2003 ) ; ( * ? ? ?
* kudoh & basu 2006 ) ) also show that the velocity dispersion is essentially alfvnic and that low density regions with high local alfvn speed support large amplitude velocity fluctuations , in agreement with the observed linewidth - size relations of molecular clouds ( ( * ? ? ?
* solomon et al . 1987 ) ) .
magnetic fields can also lower the dissipation rate of turbulence in some cases , although not necessarily dramatically ( ( * ? ? ?
* kim & basu 2013 ) ) , and allow more efficient distribution of outflow power across the cloud ( ( * ? ? ?
* wang et al . 2010 ) ) . , width=340 ] recent observations by the _ herschel space observatory _ have established that the dense cores and young stellar objects are often threaded along networks of elongated structures seen in projection ( ( * ? ? ?
* andr et al . 2010 ) ; ( * ? ? ?
* menschikov et al . 2010 ) ) .
* arzoumanian et al . 2011 ) ) .
when magnetic field morphology can be detected through polarized dust emission , the magnetic field is generally aligned perpendicular to the long axis of the filaments ( ( * ? ? ?
* palmeirim et al . 2013 ) ) .
these features are at least qualitatively consistent with the magnetic turbulent scenario described above .
another interesting result is that the measurement of average fwhm widths across filaments in three different clouds shows a cluster of values around @xmath2 pc , even though the jeans length calculated from the peak column densities are varying by about two orders of magnitude ( ( * ? ? ?
* arzoumanian et al . 2011 ) ) .
it seems that the filament widths require an explanation that goes beyond the simple jeans length prescription .
they are likely to be quasi - equilibrium objects based on the amount of fragmentation along their lengths to form dense cores .
recall that for a non - magnetized isothermal cylinder , there is a critical line mass @xmath3 ( where @xmath4 is the isothermal sound speed ) below which radial equilibrium is possible but fragmentation occurs along the axis of the cylinder .
however if the line mass exceeds this critical value , then indefinite collapse to a spindle occurs before any fragmentation can take place ( ( * ? ? ?
* inutsuka & miyama 1997 ) ) .
modified values of critical line mass , actually critical line mass to line flux ratio , for magnetized cylinders with a lateral magnetic field have recently been obtained by ( * ? ? ?
* tomisaka ( 2014 ) ) and ( * ? ? ?
* hanawa & tomisaka ( 2015 ) ) . given the fragmentation along their length ,
it is likely that the line mass to line flux ratio of observed ribbons is subcritical .
* hennebelle & andr ( 2013 ) ) developed an analytic model of filament width in which the energy input from accretion onto a filament is balanced by energy dissipation by ion - neutral friction , and there is mechanical equilibrium between gravity and internal turbulence .
they obtained a mean filament width of @xmath2 pc for a range of column densities .
an alternate model of filament formation was explored by ( * ? ? ?
* kudoh & basu ( 2014 ) ) in which a filament ( really ribbon ) is created by turbulent compression of subcritical gas and halts contraction when magnetic pressure balances ram pressure .
however , their semi - analytic model was used to estimate time scales for core formation and not to estimate filament widths .
) profile versus radius from ( * ? ? ?
* dapp et al . ( 2012 ) ) .
the thin lines are profiles in ascending order of time .
several features are identifiable via their associated breaks in the profile ( 1 ) prestellar infall profile with @xmath5 .
( 2 ) magnetic wall at @xmath6 , where the bunched - up field lines decelerate material before it continues the infall . ( 3 ) expansion wave profile with @xmath7 outside the first core .
( 4 ) first core at @xmath8 . ( 5 ) infall profile onto the second core with @xmath5 .
after the first core has reached @xmath9 , it starts to collapse , as h@xmath10 is dissociated .
( 6 ) expansion wave profile with @xmath7 outside the second core .
( 7 ) second core at @xmath11 .
[ profiles],width=340 ] scales from ( * ? ? ?
* dapp et al . ( 2012 ) ) .
the dashed lines represent the flux - freezing model , while the solid lines show _ the same _ field lines for the model including non - ideal mhd effects for a grain size @xmath12 . in both cases ,
the second core has just formed and is on the left axis midplane .
the field lines straighten out significantly on small scales in the non - ideal mhd model compared to the flux - frozen model .
[ flines],width=283 ] once a dense core is formed by fragmentation of a larger clump or filament , it will begin runaway collapse as long as its mass - to - flux ratio is supercritical .
the hydrodynamic collapse leads to a well known set of outcomes : development of an @xmath13 density profile and a shrinking central flat density peak that eventually forms the first hydrostatic core , later collapsing to the second core of stellar dimensions ( e.g. , ( * ? ? ?
* larson 1969 ) ) .
the addition of rotation and/or magnetic fields leads to a flattened collapsing core , especially at high densities .
once a central protostar is formed at the pivotal moment @xmath14 , several interesting things happen .
an expansion wave moves outward from the center ( ( * ? ? ?
* shu 1977 ) ) , changing the character of flow variables within it .
when rotation is added , a centrifugal disk is able to form , however an additional flux - frozen magnetic field leads to catastrophic magnetic braking that shuts off disk formation ( ( * ? ? ?
* allen et al . 2003 ) ; ( * ? ? ?
* galli et al . 2006 ) ) . a straightforward solution to catastrophic magnetic braking
is the presence of non - ideal mhd effects , notably ambipolar diffusion and ohmic dissipation .
these will weaken the magnetic field strength and reduce the effectiveness of magnetic braking . in (
* dapp & basu ( 2010 ) ) and ( * ? ? ?
* dapp et al . ( 2012 ) ) we used the thin - disk approximation to follow collapse to the formation of a second stellar core while including non - ideal mhd effects .
both studies show that non - ideal mhd effects allow the formation of a disk immediately after the pivotal moment @xmath14 . in (
* dapp et al . ( 2012 ) ) we used a chemical network model to calculate the ionization fraction and coefficients of ambipolar diffusion and ohmic dissipation .
figure [ profiles ] shows a series of column density profiles at various times , with the last one at the moment immediately following the formation of a second core of stellar dimensions .
each distinct region of the column density profile is captured , including the mhd accretion shock created by outward moving magnetic flux ( ( * ? ? ?
* li & mckee 1996 ) ; ( * ? ? ?
* contopoulos et al . 1998 ) ) .
figure [ flines ] shows the magnetic field lines in cases with ideal and non - ideal mhd .
the former case has extreme flared field lines and no disk is able to form due to catastrophic magnetic braking
. the non - ideal mhd model yields the development of a centrifugal disk immediately after the formation of the second core , however the loss of some angular momentum due to magnetic braking is expected to keep disks at relatively small sizes @xmath15 au in the early class 0 phase ( @xmath16 yr ) .
, log units ) at six times after the formation of the central star ( bright circle in the coordinate center ) , from ( * ? ? ?
* vorobyov & basu ( 2010a ) ) .
note the zoom - in as the time increases .
the top two rows contain images of size 600 au on each side , while the bottom row contains images of size 400 au on each side .
[ disk],width=340 ] as shown in the previous section , the collapse of a prestellar core leads to a very focused collapse with a power - law density profile . as a result
, the probability of fragmentation during the runaway collapse phase is low .
however , a centrifugally supported disk can form after a central protostar has formed .
the existence of a central point mass potential allows a centrifugal barrier to be encountered by an infalling rotating fluid element that conserved its angular momentum ( see discussion in ( * ? ? ? * basu & mouschovias 1995 ) ) .
a protostellar disk is a quasi - equilibrium structure and evolves on a time scale much longer than its dynamical time .
therefore , it has time for fragmentation if conditions are suitable , e.g. , the criteria for gravitational instability and fragmentation are satisfied ( ( * ? ? ? * toomre 1981 ) ) .
disk fragmentation is the most likely mechanism for the formation of multiple stellar systems and the formation of brown dwarfs and giant planets at wide orbits of several tens of au .
we show here the results of hydrodynamic numerical simulations of ( * ? ? ?
* vorobyov & basu ( 2010a ) ) , who modeled the formation of the disk from the collapse of a prestellar core , as well as its subsequent evolution on myr time scales .
figure [ disk ] shows images of the disk column density at a series of times after formation of the central protostar . during the first @xmath17 yr , there is vigorous episodic gravitational instability in the disk , driven by mass accretion from the parent core .
clumps form in spiral arms through nonlinear gravitational instability , and are eventually driven in to the center through gravitational interaction with trailing spiral arms . in this model ,
one clump is able to survive to late times and eventually carves out a gap in the disk and settles into a stable orbit by 4.5 myr . the clump is essentially a first hydrostatic core and is the progenitor of a low mass companion star with an orbit radius @xmath18 au .
this mechanism can also produce wide orbit proto - brown - dwarfs and giant protoplanets .
details of the numerical model and the overall scenario of the migrating embryo model can be found in ( * ? ? ?
* vorobyov & basu ( 2006 ) ) , ( * ? ? ?
* vorobyov & basu ( 2010b ) ) , and ( * ? ? ?
* vorobyov & basu ( 2015 ) ) . | i review some steps in the conversion of molecular cloud gas into stars and planets , with an emphasis in this presentation on the early stage molecular cloud fragmentation that leads to elongated filaments / ribbons .
magnetic fields can play a crucial role in all stages and need to be invoked particularly for early stage fragmentation as well as in late core collapse where it may control disk formation .
i also review some elements of hydrodynamic modeling of disk evolution . |
a topic of contemporary interest in nuclear physics is the investigation of the effective nucleon - nucleon ( nn ) interaction in a dense hadronic environment
. such environment can be produced in the laboratory via energetic heavy ion ( hi ) collisions or can be found in astrophysical systems , particularly the interior of neutron stars . in all cases predictions rely heavily on the nuclear equation of state ( eos ) , which is one of the main ingredients for transport simulations of hi collisions as well as the calculation of neutron star properties .
transport equations , such as the boltzmann - uehling - uhlenbeck ( buu ) equation , describe the evolution of a non - equilibrium gas of strongly interacting hadrons . in buu - type models ,
particles drift in the presence of the mean field while undergoing two - body collisions , which require the knowledge of in - medium two - body cross sections . in a microscopic approach , both the mean field and the binary collisions
are calculated self - consistently starting from the bare two - nucleon force .
we present microscopic predictions of nn cross sections in symmetric and asymmetric nuclear matter . in asymmetric matter
, the cross section becomes isospin dependent beyond the usual and well - known differences between the @xmath0 and the @xmath1 cases .
it depends upon the total density and the degree of proton / neutron asymmetry , which of course also implies that the @xmath2 and the @xmath3 cases will in general be different from each other . in this paper
, we will only be concerned with the the strong interaction contribution to the cross section ( coulomb contributions to the @xmath2 cross section are not considered ) .
asymmetry considerations are of particular interest at this time .
the planned rare isotope accelerator will offer the opportunity to study collisions of neutron - rich nuclei which are capable of producing extended regions of space / time where both the total nucleon density and the neutron / proton asymmetry are large .
isospin - dependent buu transport models @xcite include isospin - sensitive collision dynamics through the elementary @xmath2 , @xmath3 , and @xmath0 cross sections and the mean field ( which is now different for protons and neutrons ) .
the latter is a crucial isospin - dependent mechanism and is the focal point of an earlier paper @xcite . in a simpler approach
, the assumption is made that the transition matrix in the medium is approximately the same as the one in vacuum and that medium effects on the cross section come in only through the use of nucleon effective masses in the phase space factors @xcite . concerning microscopic calculations ,
earlier predictions can be found , for instance , in refs .
@xcite , but asymmetry considerations are not included in those predictions . on the other hand , it is important to investigate to which extent the in - medium cross sections are sensitive to changes in proton / neutron ratio , the main purpose of this paper .
in - medium cross sections are necessary to study the mean free path of nucleons in nuclear matter and thus nuclear transparency .
the latter is obviously related to the total reaction cross section of a nucleus , which , in turn , can be used to extract nuclear r.m.s .
radii within glauber - type models @xcite .
therefore , accurate in - medium _ isospin - dependent _ nn cross sections can ultimately be very valuable to obtain information about the size of exotic , neutron - rich nuclei . in the next section , after providing some details on the calculation , we present and discuss our results .
our conclusions are summarized in section iii .
details of our application of the dirac - brueckner - hartree - fock framework to asymmetric matter can be found in ref.@xcite . we choose the bonn - b potential @xcite as our model for the free - space two - nucleon ( 2n ) force . the nuclear matter calculation of ref .
@xcite provides , together with the eos , the single - proton / neutron potentials as well as their parametrization in terms of effective masses @xcite .
those effective masses , together with the appropriate pauli operator ( depending on the type of nucleons involved ) , are then used in a separate calculation of the in - medium reaction matrix ( or @xmath5-matrix ) under the desired kinematical conditions .
our calculation is controlled by the total density , @xmath6 , and the degree of asymmetry , @xmath7 , with @xmath8 and @xmath9 the neutron and proton densities . for the case of identical nucleons ,
the @xmath5-matrix is calculated using the appropriate effective mass , @xmath10 , and the appropriate pauli operator , @xmath11 , depending on @xmath12 , where @xmath13 or @xmath14 . for non - identical nucleons
, we use the `` asymmetric '' pauli operator , @xmath15 , depending on both @xmath16 and @xmath17 @xcite . in fig . 1
, we show the variations of @xmath16 and @xmath17 with increasing neutron fraction at three fixed densities , according to the relations @xmath18 @xmath19 this may facilitate the interpretation of results later on . in the usual free - space scattering scenario ,
the cross section is typically represented as a function of the incident laboratory energy , which is uniquely related to the nucleon momentum in the two - body c.m .
frame , @xmath20 ( also equal to the relative momentum of the two nucleons ) , through the well - known formula @xmath21 . in nuclear matter , though , the pauli operator depends also on the total momentum of the two nucleons in the nuclear matter rest frame .
this could be defined in some average ( density - dependent ) manner , or kept as an extra degree of freedom on which the cross section will depend .
we will take the latter approach and examine this aspect in the next subsections . another issue to consider carefully when approaching the concept of in - medium cross sections is the non - unitary nature of the interaction .
due to the presence of pauli blocking , the in - medium scattering matrix does not obey the free - space unitarity relations through which phase parameters are usually defined and from which it is customary to determine the nn scattering observables . as done in ref .
@xcite but unlike ref .
@xcite , we calculate the cross section directly from the scattering matrix elements thus avoiding the assumption of in - vacuum unitarity and its consequences ( such as the optical theorem ) .
that is , we integrate the differential cross section @xmath22 where @xmath23 contains the usual sum of amplitudes squared and phase space factors .
an alternative way would be to include pauli blocking in the definition of phase parameters @xcite . as a baseline , we start by showing the @xmath2 cross section as a function of the momentum @xmath20 and at different densities of symmetric matter , see fig . 2 . until otherwise noted , we use in - vacuum kinematics to define the total 2n momentum in the nuclear matter rest frame ( that is , the target nucleon is , on the average , at rest ) .
the range of momentum in fig . 2 corresponds to values of the in - vacuum laboratory energy between approximately 20 and 260 mev . as compared to the predictions of ref .
@xcite , where the use of the @xmath24-matrix approach imposes unitarity of the in - medium scattering matrix , the suppression of the cross section with increasing density is less pronounced .
this was already pointed out in ref .
@xcite and is most likely due to the different handling of the unitarity issue as noted above .
overall , we find reasonable qualitative agreement with previous predictions .
we also notice that the high - density cross section tends to rise again at large momenta , a behavior already observed in refs .
@xcite particularly for the @xmath2 channel
. a simpler approach to calculating in - medium cross sections consists of scaling the free - space value according to the relation @xcite @xmath25 where @xmath26 is the effective mass . in fig .
3 we compare this approximation with the predictions from our full calculation .
we see that at low momentum , and particularly at the lower densities , the agreement is quite good , but it strongly deteriorates at higher momenta and densities .
medium effects on the interaction are clearly important , particularly the interplay between the ( energy - dependent ) pauli blocking mechanism and the dressing of the quasiparticle due to the surrounding medium .
we next examine the role of isospin asymmetry , the focal point of this paper . in fig . 4 ,
the ratio of the in - medium cross section to the free - space one is shown for both @xmath2 and @xmath3 scatterings as a function of the asymmetry .
the total density is fixed to its value near saturation and the momentum @xmath20 is chosen to be equal to the corresponding fermi momentum ( @xmath27=1.3 @xmath28 ) .
the dependence on @xmath29 is generally weak , which may be attributed to the `` competing '' roles of the effective mass and pauli blocking . for the @xmath2 case , for instance , the smaller effective mass @xcite and the lower fermi momentum ( see fig .
1 ) would tend to decrease and increase the cross section , respectively .
the opposite happens in the @xmath3 case .
for the momentum considered in fig . 4 , the role of the effective mass appears to be dominant , lowering the @xmath2 cross section and raising the @xmath3 one .
this trend of the @xmath2/@xmath3 cross sections versus @xmath29 and relative to each other is in qualitative agreement with predictions based on scaling the cross section as in eq.(4 ) , with effective masses obtained from the modified gogny interation recently used in isospin - dependent buu calculations @xcite .
figure 5 displays the cross section ratios for fixed values of the asymmetry and density but changing relative momentum .
there , we see that the ratio approachs one in the high - momentum limit . in fig .
6 , the cross section ratios are shown as functions of the ( average ) fermi momentum but fixed asymmetry and @xmath20 .
the latter is chosen as in fig .
the cross sections are seen to go down rather quickly with increasing density but then rise again at the higher densities . as we mentioned earlier ,
this was already observed , particularly in the @xmath2 channel , in previous microscopic calculations of cross sections in symmetric matter @xcite . in this region of the phase space , disagreement with predictions based on eq .
( 4 ) @xcite reflects the discrepancy already noted in the high - density part of the lower panel in fig . 3 .
following up on the comments we made at the end of subsection iia , in the next figures we examine the impact of treating the 2n total momentum with respect to nuclear matter , @xmath30 , as an independent variable ( rather than linked to @xmath20 through the simplifying assumption of in - vacuum kinematics ) . in fig . 7 , we show @xmath2 and @xmath3 cross section ratios under the same conditions of density and two - nucleon relative momentum as in fig . 4 but changing value of the total 2n momentum , @xmath30 .
the general tendency seems rather independent of @xmath30 , although some minor differences can be seen with fig . 4 as well as among the three situations displayed in fig .
in particular , we notice that for the lower values of @xmath30 the @xmath2 cross section shows a lesser tendency to decrease with increasing @xmath29 , and perhaps even a slight tendency to rise again .
this may be understood in terms of pauli blocking , which becomes more important at lower values of @xmath30 and whose effect is opposite the one coming from the decreasing proton mass .
next , to broaden our previous analysis , we show in fig . 8 and fig .
9 situations analogous to those presented in fig . 4 and fig . 7 , respectively , but at a lower density .
the chosen value of the average fermi momentum , which is also taken as the value of @xmath20 , corresponds to approximately one - half of saturation density .
sensitivity to the asymmetry appears stronger at lower density , as seen from the more pronounced differences between the @xmath2 and @xmath3 curves in both fig . 8 and fig .
similarly to what was done in fig . 2 for the @xmath2 case , we begin with showing in fig . 10 the basic density / momentum dependence of the @xmath0 cross section in symmetric matter .
( in - vacuum kinematics is used at this point . )
we notice here that the cross section at the density equivalent to @xmath27=1.1 @xmath28 shows some local enhancement which is considerably more pronounced than what can be seen in the @xmath2 case , see fig .
this comparison suggests that enhanced attraction is generated at low density in the t=0 partial waves , most likely those with the quantum numbers of deuteron . in fig . 11 , we explore more closely the dependence of these structures " on the kinematics
the label `` fsk '' stands for free - space kinematics ( as was used in fig .
10 ) , whereas the other three curves refer to different choices of the total momentum @xmath30 as indicated in the figure .
clearly , the location of these enhancements is determined by the kinematics , moving to lower values of @xmath20 for higher values of @xmath30 .
although noticible , these gentle fluctuations do not resemble at all the extremely large resonance - like peaks reported in some earlier works @xcite and interpreted as precursor of a superfluid phase transition in nuclear matter at particular densities / momenta .
although crucial for generating those resonances in ref .
@xcite was the inclusion of hole - hole scattering , other authors have reported large bound - state signatures even without hole - hole contributions @xcite .
our predictions do not confirm those findings .
very mild fluctuations are predicted in ref .
@xcite , whereas no enhancements at all appear in the calculations of ref .
@xcite .
we will focus next on the asymmetry dependence of the @xmath0 cross section . having observed in the previous subsection that asymmetry effects on protons and neutrons tend to move the cross section in opposite directions
, we may reasonably expect very mild dependence on @xmath29 .
this is indeed confirmed in fig .
12 ( where in - vacuum kinematics is used ) , and fig . 13 ( where , as before , the sensitivity to the choice of @xmath30 is explored ) . for in - medium @xmath0 scattering ,
the nature of the cross section is the result of a subtle combination of multiple effects .
these come from both the neutron and proton effective masses and the role of both the neutron and proton fermi momenta in the expression of the ( asymmetric ) pauli operator @xcite . in the end , the behavior versus @xmath29 is nearly flat , although a very slight tendency to rise can be identified , with the exception of the lower-@xmath30 curve in fig .
we have presented microscopic calculations of cross sections for scattering of nucleons in neutron - rich matter .
the sensitivity to the asymmetry in neutron / proton ratio comes through the combined effect of pauli blocking and changing effective masses .
the lowering(rising ) of the proton(neutron ) fermi momentum and the reduced(increased ) proton(neutron ) effective mass tend to move the cross section in opposite directions . in the phase space regions we have examined here ,
the general tendency is for the @xmath2 cross section to go down and the @xmath3 one to rise with increasing asymmetry .
this indicates that the role of the effective mass is dominant over pauli blocking . when both kinds of nucleons are involved , though , dispersive and pauli blocking effects on each type of nucleon result in large cancelations .
the @xmath0 cross section is essentially @xmath29-independent , although minor variations can be seen depending on the kinematics .
on the other hand , the basic density / momentum dependence of the @xmath0 cross section in symmetric matter appears rather interesting .
different predictions disagree concerning the presence of large bound - state effects which might be associated with the onset of superfluidity . in summary , sensitivity to the asymmetry can be non - negligible for scattering of identical nucleons .
the degree of sensitivity depends on the region of the energy - density - asymmetry phase space under consideration .
we conclude that the proton mean free path could be affected in a significant way by in - medium scattering , mostly through @xmath31 . to test these findings
, it is very useful to identify hi collision observables specifically sensitive to the two - body cross sections .
possible probes of the isospin - dependence of the in - medium nn cross sections are presently being investigated @xcite . whereas traditional means such as measurements of the stopping power are found to be ambiguous with respect to determining isospin dependence , potential candidates include isospin tracers like the neutron / proton free - nucleon ratio in reactions induced by radioactive beams in inverse kinematics @xcite .
+ + see , for instance , b.a .
li , phys .
85 * , 4221 ( 2000 ) , and references therein .
f. sammarruca , w. barredo , and p. krastev , phys .
c * 71 * , 064306 ( 2005 ) .
v.r . pandharipande and s.c .
pieper , phys .
c * 45 * , 791 ( 1992 ) .
d. persram and c. gale , phys .
c * 65 * , 064611 ( 2002 ) .
li and r. machleidt , phys .
c * 48 * , 1702 ( 1993 ) ; * 49 * , 566 ( 1994 ) . c. fuchs , a. faessler , and m. el - shabshiry , phys .
c * 64 * , 024003 ( 2001 ) .
see , for instance , r. crespo and r.c .
johnson , phys .
c * 60 * , 034007 , ( 1999 ) , and references therein .
d. alonso and f. sammarruca , phys .
c * 67 * , 054301 ( 2003 ) .
r. machleidt , adv .
* 19 * , 189 ( 1989 ) .
t. alm , g. ropke , and m. schmidt , phys .
c * 50 * , 31 ( 1994 ) .
bao - an li , private communications .
a. bohnet , n. ohtsuka , j. aichelin , r. linden , and a. faessler , nucl . phys . *
a494 * , 349 ( 1989 ) .
li , p. danielewicz , and w. lynch , nucl - th/0503038 . | we calculate nucleon - nucleon cross sections in the nuclear medium with unequal densities of protons and neutrons .
we use the dirac - brueckner - hartree - fock approach together with realistic nucleon - nucleon potentials .
we examine the effect of neutron / proton asymmetry and find that , although generally mild , it can be significant in specific regions of the phase space under consideration . |
dwarf stars with subsolar metallicity are bluer than solar abundance dwarfs or main - sequence stars of the equivalence mass .
they lie below the main sequence in the hertzsprung russell diagram and appear less luminous than main - sequence stars .
these objects were thus called subdwarfs " by @xcite .
evolving subdwarfs are referred to as cool subdwarfs to provide distinction from hot subdwarfs , a different class of objects ( e.g. * ? ? ?
cool dwarfs and cool subdwarfs are observationally and kinematically distinct , and were separated into population i and ii categories , respectively by @xcite .
population i and ii are associated with the galactic disc and spheroid respectively .
@xcite find that the old , high - velocity population ii stars were also metal deficient .
red dwarfs are low - mass and relatively cool stars on the main sequence with spectral types of late - type k and m , masses between @xmath0 0.6 and @xmath0 0.08 @xmath3 , and surface effective temperatures between @xmath0 4000 and @xmath4 2300 k @xcite .
red dwarfs are the most common type of star both in the milky way @xcite and other galaxies @xcite .
the spectra of red dwarfs are dominated by molecular absorption bands of metal oxides and hydrides .
tio and cah near 7000 are the most prominent bands @xcite .
red subdwarfs ( rsds ) are the subsolar counterparts of red dwarfs by chemical abundance .
rsds are significantly rarer than red dwarfs .
the kapteyn s star ( sdm1 type ) is the only cool subdwarf among the sample of 8 pc within the sun which contains more than 244 known stars and brown dwarfs , including 157 m dwarfs @xcite .
the known sample of late - type m subdwarfs is significantly smaller than that of early - type m subdwarfs because they are fainter and have lower space density according to the halo mass function ( e.g. * ? ? ?
late - type m subdwarfs have the most complex stellar atmospheres because they are ultracool and have large - scale variation of chemical abundance and gravity . although rsds are less luminous than f , g and mid - k subdwarfs , they are numerous and have more notable spectral features caused by chemical abundance and gravity , thus they are better targets for observational and theoretical studies .
it is difficult to distinguish f , g and mid - k subdwarfs from normal dwarf stars of the same spectral type using their optical spectra because they are featureless .
dwarfs and subdwarfs with spectral types of late - type k and m , however , have a number of very different spectral features .
model atmospheres also suggest that optical spectra of m subdwarfs are dramatically affected by metallicity variations .
it is thus possible to use low - resolution spectroscopy for exploring effective temperature , metallicity and gravity effects on the spectra of these cool objects @xcite .
classification and characterization of m subdwarfs is a rapidly evolving field @xcite .
m subdwarfs are classified into three metal classes : subdwarf ( sdm ) , extreme subdwarf ( esdm ) and ultra subdwarf ( usdm ) based on the ratio of tio to cah indices @xcite .
cah and tio indices are easy to measure and sensitive to temperature and metallicity .
however , @xcite found that the cah and tio indices are affected in complicated ways by combinations of temperatures , metallicities and gravities of rsds .
model spectra show that the tio5 index is more sensitive to metallicity while the cah2 and cah3 indices are more sensitive to gravity .
this suggests that the effect of gravity , which was previously ignored , should be considered in the classification of m subdwarfs .
an ideal testbed for the impact of gravity on spectra of m subdwarfs is binary systems with two m subdwarfs which share the same age and metallicity . with the same effect on spectra from metallicity , it is possible to measure the difference of gravity on the broad indices of cah2 , cah3 and tio5 .
m subdwarfs in binary systems are the key for both model calibration , and spectral classification and characterization .
discovery of a sample of rsd binary systems is therefore crucial . in this paper
we present the discovery of 45 rsd binary systems from the sloan digital sky survey ( sdss , * ? ? ?
* ) and the ukirt infrared deep sky survey ( ukidss , * ? ? ?
at least one companion in each system is confirmed as an rsd with sdss spectra .
the selection and classification processes of our rsd sample are presented in section [ sample ] .
the identification of rsd binary systems is presented in section [ sbinary ] .
section [ snote ] presents further discussion of rsd binary systems of particular interests .
summary and conclusions is described in section [ ssummary ] .
the eighth data release ( dr8 ) of sdss includes 14555 deg@xmath5 of imaging data , and 9274 deg@xmath5 of spectroscopic data .
there are over 1.84 million spectra in total , including 0.6 million stars , 0.13 million quasars and 0.95 million galaxies @xcite .
the sdss dr8 also includes proper motions ( pms ) for objects derived by combining sdss astrometry with usno - b positions , re - calibrated against sdss .
the errors of pms are typically less than 10 mas@xmath6yr@xmath7 @xcite .
band reduced pms @xmath8 and @xmath9 colour of our pm selected objects .
symbols in the figure are black down - pointing triangles : m dwarfs ; green squares : sdms ; red circles : esdms ; blue up - pointing triangles : usdms ; grey left - pointing triangles : wds . ]
we selected our candidates using sdss casjobs by combining the spectroscopic and pm catalogues .
we required pm greater than 100 mas@xmath6yr@xmath7 .
no photometric criteria were applied but spectral observations for red dwarfs in sdss is limited to @xmath10 and @xmath11 .
most of rsds in sdss spectroscopic data base are at distances of 200 @xmath0 400 pc .
objects with tangential velocity of 200 km s@xmath7 at a distance of 400 pc ( or 100 km s@xmath7 and 200 pc ) will have pms higher than 100 mas@xmath6yr@xmath7 .
a pm cut of 100 mas@xmath6yr@xmath7 allows a good balance in order to select most late - type k , early - type m and almost all mid - late type m subdwarfs in the combined sdss pm + spectroscopic catalogues while minimizing the contamination by dwarf stars . since not all spectra in dr7 @xcite
are reproduced in dr8 we applied our search to both data releases finding 7499 and 8445 spectra respectively .
there were 9146 spectra in total for 8236 objects ( some objects had more than one spectrum ) .
sdss spectra were reduced and classified with the idlspec2d code by sdss @xcite . for 8236 objects in our sample ,
5687 were classified as stars , 1558 as galaxies and 348 as qsos , 643 of them were not classified .
not all objects in the sample are classified by sdss .
rsds are generally classified as stars or galaxies in some cases . to pick out and classify rsds in our sample properly we used a k and m subdwarf classification code developed by @xcite .
we ran the code on the spectra of all 8236 objects to identify subdwarfs and assign their spectral types .
the code classified objects into nine groups : dm ( 2455 ) , dk ( 80 ) , sdm ( 689 ) , sdk ( 326 ) , esdm ( 483 ) , esdk ( 189 ) , usdm ( 256 ) , usdk ( 189 ) , and unclassified ( 3442 ) .
we found many objects , originally classified as galaxies by sdss , that were re - classified as sdk , esdk or usdk .
we inspected the spectra by eye in each group to ensure that the correct classification was applied in each case .
we found 2004 galaxies with false pms were selected into our sample : 383 of them were classified as stars ( mostly late - type k subdwarfs ) .
we found that 463 late - type k subdwarfs and 1363 m subdwarfs survived the eyeball check .
54 objects were classified as sdm but removed from our subdwarf sample because they do not have typical halo kinematics .
[ hr ] shows reduced pms and @xmath9 colour of these objects .
three sequences from left to right show the location of white dwarfs ( wds ) , m subdwarfs and m dwarfs .
[ ugrz ] shows two colour diagrams of @xmath12 versus @xmath9 and @xmath12 versus @xmath9 of the sample .
the wd sample will be discussed in a future paper . | red subdwarfs in binary systems are crucial for both model calibration and spectral classification .
we search for red subdwarfs in binary systems from a sample of high proper motion objects with sloan digital sky survey spectroscopy .
we present here discoveries from this search , as well as highlight several additional objects of interest .
we find 30 red subdwarfs in wide binary systems including : two with spectral type of esdm5.5 , 6 companions to white dwarfs and 3 carbon enhanced red subdwarfs with normal red subdwarf companions .
15 red subdwarfs in our sample are partially resolved close binary systems . with this binary sample
, we estimate the low limit of the red subdwarf binary fraction of @xmath0 10 % .
we find that the binary fraction goes down with decreasing masses and metallicities of red subdwarfs .
a spectroscopic esdk7 subdwarf + white dwarf binary candidate is also reported .
30 new m subdwarfs have spectral type of @xmath1m6 in our sample .
we also derive relationships between spectral types and absolute magnitudes in the optical and near - infrared for m and l subdwarfs , and we present an m subdwarf sample with measured @xmath2 space velocities .
[ firstpage ] brown dwarfs stars : carbon stars : late - type stars : population ii subdwarfs
galaxy : halo . |
vacuum fluctuations are an outstanding consequence of the quantum theory of the electromagnetic radiation field , and observable manifestations of vacuum fluctuations include radiative level shifts and casimir forces @xcite .
these forces are long - range interactions between neutral objects due to their interaction with vacuum fluctuations ; they are quantum effects of the radiation field and have not a classical analogue .
usually casimir effect refers with this kind of interactions between macroscopic objects @xcite , and casimir - polder forces with the interactions of atoms or molecules with a macroscopic object such as a surface or among atoms and molecules @xcite .
vacuum fluctuations are changed by the presence of boundary conditions given by dielectric or metallic surfaces .
this change of electric and magnetic field fluctuations gives rise to energy shifts of atoms or molecules , placed near the surface , that depend from their position , yielding casimir - polder forces @xcite .
evaluating field fluctuations around a boundary allows also to obtain how interactomic casimir - polder forces are modified by the presence of the boundary @xcite .
these facts give further elements making relevant studying the dependence of vacuum fluctuations or vacuum energy densities on the distance from a dielectric or conducting surface . in the case of a perfectly conducting infinite plate , the vacuum electric and magnetic fluctuations @xmath0 and @xmath1 , after subtraction of the homogeneous zero - point terms existing even in the absence of the plate ,
behave as @xmath2 , @xmath3 being the distance from the plate ; they do diverge at the metal - vacuum interface .
the physical origin of this divergence and the possible presence of further singular terms at the interface has been recently questioned in the literature @xcite , and possibility of their regularizations has been investigated in the case of a scalar field using an appropriate potential to represent the wall @xcite . it has been shown that , in the case of a half - space filled with a non - dissipative dielectric material characterized by a real dielectric constant @xmath4 independent from the frequency , such divergences at the interface are still present in the limit @xmath5 @xcite .
an analogous behavior has been obtained also if dispersion is introduced using the plasma model @xcite . in the case of a scalar field , it has been shown that the stress - tensor components are regularized by reflection and transmission coefficients ; also , the surface divergence associated to a perfectly reflecting mirror is canceled by a new energy density near the surface @xcite .
surface divergences in the casimir energies has also been considered by introducing extra terms in the lagrangian in the form of a @xmath6-function potential , in order to simulate the boundary condition @xcite .
research in this direction has been also motivated by the quest for situations where negative energy densities occur @xcite , and solving discrepancies between total self - energies and the local energy densities @xcite .
the presence of surface divergences may be relevant also from the point of view of the coupling of the related field energy densities with gravity @xcite . in this paper we shall consider electric and magnetic field fluctuations in the vacuum region near a half - space filled with a homogeneous nondissipative dielectric characterized by a real refractive index @xmath7 .
the field fluctuations are evaluated using the carniglia - mandel triplets as field modes .
a time - splitting procedure is used in evaluating the frequency integrals , and we notice that this is mathematically equivalent to introduce an exponential cut - off function @xmath8 , with @xmath9 playing the role of a cut - off frequency .
we then consider the limit @xmath10 , leading to a model for a metallic material .
if the quantity @xmath11 introduced by the time - splitting procedure is kept finite , this is representative of a situation where the contribution of field modes with @xmath12 is suppressed by the exponential cut - off .
we can then consider the limit of an ideal metallic plate by taking the limit @xmath13 .
after this limit , the cut - off function is one for all frequencies and we recover the nondispersive case . with this procedure
, the ideal metal case is obtained through a limiting procedure , and at no point we need to use boundary conditions for ideal metals .
we find that in the limit @xmath10 the renormalized field fluctuations are finite for any finite value of @xmath14 .
we are also able to show that in the ideal metal limit @xmath15 surface divergences at the metal - vacuum interface emerge , and to investigate their properties . as far as we know
, the structure of these surface divergences can not be derived if the field is directly quantized with a perfectly conducting boundary .
moreover , no discrepancy between total self - energy and local energy densities is present in our model .
such discrepancies indeed occur when energy densities are evaluated directly for the ideal metallic plate , because of the singular behavior at @xmath16 . from the field energy density we can also evaluate the electric and magnetic casimir - polder force on an atom placed near the dielectric half - space
. strong radiative interactions of cesium atoms near dielectric boundaries have been recently investigated experimentally @xcite . in sec .
[ sec:2 ] we calculate the vacuum fluctuations of the electric and magnetic field in the vacuum space in presence of a nondissipative homogeneous dielectric half - space .
we use quantization in terms of the carniglia - mandel triplets , that include evanescent waves , and we use the known time - splitting procedure . in sec .
[ sec:3 ] we consider the ideal metal case by a limit procedure from the dielectric case , and we analyze in detail the field fluctuations ( and field energy densities ) , in particular in proximity of the metal - vacuum interface .
the emergence and structure of surface divergences at the interface in the ideal conductor limit is discussed in detail .
we consider an half - space filled with an homogeneous dielectric medium and the vacuum in the other half - space .
we label @xmath3 the direction orthogonal to the dielectric - vacuum interface : @xmath17 is the dielectric half - space and @xmath18 the vacuum half - space .
our model has a translational symmetry in the x - y directions .
we assume that the dielectric is non - dissipative with a ( real ) dielectric constant independent on the frequency .
we shall later discuss how to introduce in our model a dependence on the frequency , in particular when considering the limit for an ideal conductor .
first step is the calculation of the square of the electric and magnetic fields in the vacuum half - space ( @xmath18 ) in the ground state of the field .
we use the well - known quantization scheme in terms of the carniglia - mandel triplets @xcite .
the carniglia - mandel modes are given by the following expressions [ tripletticm ] @xmath19 @xmath20 @xmath21 @xmath22 where the index @xmath23 indicates waves propagating toward the dielectric - vacuum interface from left or right , respectively , and @xmath24 specifies transverse electric ( te ) modes and @xmath25 transverse magnetic ( tm ) modes .
we have also defined the polarization operators @xmath26 where @xmath27 and @xmath28 .
these operators , acting on the plane - wave parts of , give the appropriate polarization unit vectors . in order to distinguish waves propagating in the positive and negative directions of the @xmath3 axis , we have defined in the free space ( @xmath18 ) the wavevectors @xmath29 inside the dielectric ( @xmath17 )
the wavevectors are @xmath30 with @xmath31 at the dielectric - vacuum interface ( @xmath16 ) , @xmath32 and the frequency @xmath33 are conserved , while @xmath34 appropriately changes in order to ensure the required continuity of @xmath35 , @xmath36 and @xmath37 . for @xmath38 less than @xmath39
, @xmath40 becomes imaginary and a total internal reflection of l - modes occurs at the interface yielding an evanescent wave in the vacuum region . thus evanescent waves are correctly taken into account by using the carniglia - mandel modes .
proof of orthogonality and completeness of these modes can be found in @xcite .
similar modes have recently been used for the quantization of the electromagnetic field in the presence of a nondispersive and nondissipative dielectric slab @xcite .
the explicit expression of the modes for @xmath18 ( vacuum region ) are [ triplettizetapositivo ] @xmath41;\ ] ] @xmath42 ; \end{aligned}\ ] ] @xmath43 @xmath44 in terms of the modes , we can write the expressions of the electric and magnetic field operators @xmath45\\ + \sum_j\int_{{k_{dz}}>0}\!\!\!\!\!\!\!\!\!d^3\mathbf{k}_d\ i\sqrt{2\pi\hbar\omega_k}\ \biggl[{b_{\mathbf{k}_dj}}\ e^{-i\omega_k t}\ { \mathbf{f}_l\bigl(\mathbf{k}_dj,\mathbf{r}\bigr)}- { b_{\mathbf{k}_dj}^\dag}e^{i\omega_k t}\ { \mathbf{f}^*_l\bigl(\mathbf{k}_dj,\mathbf{r}\bigr)}\biggr ] ; \end{split}\ ] ] @xmath46\\ + \sum_j\int_{{k_{dz}}>0}\!\!\!\!\!\!\!\!\!d^3\mathbf{k}_d\ \sqrt{\frac{2\pi\hbar c^2}{\omega_k}}\ \biggl[{b_{\mathbf{k}_dj}}\ e^{-i\omega_k t}\ \nabla\times{\mathbf{f}_l\bigl(\mathbf{k}_dj,\mathbf{r}\bigr)}+ { b_{\mathbf{k}_dj}^\dag}e^{i\omega_k t}\ \nabla\times{\mathbf{f}^*_l\bigl(\mathbf{k}_dj,\mathbf{r}\bigr)}\biggr ] . \end{split}\ ] ] ( operators @xmath47 , @xmath48 and @xmath49 , @xmath50 refer to r- and l - modes , respectively ) . the field hamiltonian , after subtraction of the zero - point energy , is given by @xmath51 where the annihilation and creation operators for photons in the different field modes satisfy the usual bosonic commutation rules @xmath52=&\ \delta_{jj'}\ \delta^3(\mathbf{k}-\mathbf{k}')\\ \bigl[{b_{\mathbf{k}_dj}},{b_{\mathbf{k}_d'j'}^\dag}\bigr]=&\
\delta_{jj'}\ \delta^3(\mathbf{k}_d-\mathbf{k}_d')\\ \end{aligned}\ ] ] ( all other commutators vanish ) . using the expression for the electric field with the modes , we can evaluate the average value of a space - time correlation of the electric field in the vacuum space near the interface @xmath53 on the ground state of the field , obtaining ( this calculations is similar to that in @xcite , and we are giving some detail of it in order to show how the time - splitting procedure we are going to use is equivalent to a high - frequency cut - off ) @xmath54\\ \end{aligned}\ ] ] with @xmath55 .
this quantity diverges for @xmath56 and @xmath57 , but we can use a point - splitting procedure , by introducing the following quantity @xmath58 with @xmath59 . in the next section
we shall show that this time - splitting procedure is mathematically equivalent to introduce a cut - off frequency in the frequency integrals , and this allows us to obtain the ideal conductor case as a limit process .
thus , for @xmath60 we have @xmath61 .
as long as @xmath62 is finite , the integrals do not diverge and we obtain @xmath63 in this expression , contributions from both traveling and evanescent waves are taken into account . after some algebraic calculations , using and , we obtain the following integral expression for @xmath64 for @xmath18 and arbitrary @xmath7 @xmath65\\ & \left .
+ \frac{|{k_z}|}{{k_{dz } } } \left[\frac{4{k_{dz}}^2}{{k_{dz}}^2+|{k_z}|^2 } + \frac{4n^2{k_{dz}}^2}{{k_{dz}}^2+n^4|{k_z}|^2}\right ] e^{-2|{k_z}|z}\right\ } e^{-\eta ck}. \end{aligned}\ ] ] an analogous expression can be obtained for the magnetic part ( @xmath18 and arbitrary @xmath7 ) @xmath66\\ & \left .
+ \frac{|{k_z}|}{{k_{dz } } } \left[\frac{4{k_{dz}}^2}{{k_{dz}}^2+|{k_z}|^2 } + \frac{4n^2{k_{dz}}^2}{{k_{dz}}^2+n^4|{k_z}|^2}\right ] e^{-2|{k_z}|z}\right\ } e^{-\eta ck}. \end{aligned}\ ] ] in the next section we shall consider the appropriate limit of these expressions for the case of an ideal metal .
we now consider the limits @xmath67 and @xmath10 of , respectively yielding the case of vacuum space and of the conductor , @xmath68 @xmath69 analogous expressions are obtained from for the magnetic fluctuations , for the vacuum space and for the conductor case , @xmath70 @xmath71 the conductor results , both for the electric and magnetic components , can be renormalized by subtracting the spatially homogeneous vacuum contribution , that is that obtained respectively in and in the absence of the material half - space .
thus we obtain @xmath72 and @xmath73 from the expressions above , it is evident that the time - splitting procedure introduces an exponential cut - off in the frequency integrals , suppressing contributions of modes with frequency @xmath74 .
if @xmath14 in kept finite , the renormalized field fluctuations and are finite for any @xmath3 , including at the interface @xmath16 , due to the regularization introduced by the cut - off function .
also , the electric and magnetic parts are opposite each other in all points and for any value of @xmath62 .
if we take @xmath75 , we get a divergence at the conductor - vacuum interface @xmath16 : in this case , a cut - off function such as that given by the plasma model for a real conductor ( assumed valid also at very high frequecies ) , that introduces a @xmath76 factor for frequencies larger than the conductor plasma frequency , would not be enough to avoid the divergence at the interface , as showed in @xcite .
there is however some controversy in the literature about which model of a real conductor ( plasma , drude or other ) should be used for calculating casimir energies and the casimir force between real materials .
an explicit evaluation of the integrals in , , yields : @xmath77 for the vacuum case ; @xmath78 for the conductor case , and @xmath79 for the renormalized conductor case .
as expected , @xmath80 and @xmath81 diverge for @xmath82 .
the limit @xmath82 of for @xmath83 gives back the well - known result for the vacuum average value of the electric field squared in the presence of a perfectly conducting plate @xcite @xmath84 analogous results are obtained for the magnetic field fluctuations @xmath85
@xmath86 @xmath87 @xmath88 the renormalized field fluctuations @xmath89 and @xmath90 given by and , which are proportional to the renormalized electric and magnetic energy densities , are finite provided @xmath62 is not vanishing .
equations and show also , as it is already known , that the presence of a perfectly conducting plate increases the fluctuations of the electric field whereas reduces magnetic field fluctuations .
we can now address the main point of this paper , that is the limit to an ideal conductor , and in particular the behavior of the field fluctuations near the plate , and the presence and origin of surface divergences in our limit to the ideal conductor .
as already noticed , the use of the time - splitting procedure introduces an exponential cut - off function in the frequency integrals of sec.[sec:2 ] giving field fluctuations .
this is equivalent to assume a cut - off function with the scale of the exponential related to a frequency such that modes of higher frequency are suppressed .
a similar cut - off function has already been used in @xcite and in @xcite for the scalar field case .
thus we can take expression and as the renormalized fluctuations of the electric and magnetic field for this model of a conductor with an exponential cut - off function . in this model
, @xmath11 plays a role analogous to that of a plasma frequency ( but we are not dealing with the plasma model , that yields a different form of the cut - off function ) . the fact that an exponential cut - off could be not representative of a real conductor is not essential in our case , because our interest is to consider the limit @xmath82 or @xmath15 : in this limit our cut - off function is approaching one for all frequency , and we recover the physical situation of a nondispersive material .
however , our procedure of considering the ideal conductor through a limiting process allows us to show the existence of surface divergences , and discuss their structure .
as we shall show in the following , this does not seems possible if the field is directly quantized with the perfectly conducting plate , nor if the plasma model is used in the limit of a large plasma frequency ( because , as it is shown in @xcite , the plasma model does not completely remove the divergences at the metal - vacuum interface ) .
we first notice that , for any @xmath59 , the integrals of the renormalized energy densities over all vacuum space vanish for both their electric and magnetic parts , because @xmath91 thus both electric and magnetic renormalized energy densities vanish when integrated over the @xmath18 half - space , whichever the cut - off frequency is .
this also means that , in the ideal conductor limit @xmath82 , the ( diverging ) electric and magnetic energy for @xmath18 , proportional to the spatial integrals of and respectively ( both behaving as @xmath2 for @xmath92 ) , must be canceled by electric and magnetic energies confined at @xmath16 in the form of surface divergences at the conductor - vacuum interface .
the existence of surface divergences was already guessed in @xcite in the case of an ideal flat boundary condition for a massless scalar field ; additional @xmath93function terms were added in the lagrangian in order to take into account the presence of the plate .
a similar situation occurs for a scalar field with appropriate diverging potentials which simulate the boundary @xcite .
our approach of obtaining the ideal conductor through an appropriate limit process , as we will now show , actually allows us to physically understand the origin and properties of such surface divergences of the electric and magnetic energy densities , without additional hypothesis .
we now consider the behavior of the electric fluctuations and energy density for growing values of @xmath11 , and compare it with the same quantity for an ideal metal as given by .
figure [ figdivergences ] shows @xmath94 as given by with a value of @xmath11 comparable with a typical plasma frequency of a real metal , in comparison with the @xmath2 behavior of @xmath95 given by .
significant differences are evident in the proximity of the interface . with @xmath96hz (
dashed blue line ) and @xmath95 ( continuous black line).,width=321 ] for different values of @xmath62 .
when @xmath62 decreases ( @xmath14 increases ) , renormalized field fluctuations and energy densities for @xmath18 tend to the @xmath2 law of the ideal conductor case ( continuous line ) , whereas at @xmath92 surface divergences appear with a positive and a negative peak that in the limit tend to squeeze at @xmath16.,width=328 ] figure [ figcomdivergences ] shows the behavior of @xmath94 for different values of @xmath62 , and compared with @xmath95 ( continuous line ) . for any nonvanishing value of @xmath62 , i.e. for any finite value of the cut - ff frequency @xmath14 , renormalized fluctuations and energy densities are finite in all points of the vacuum region and there are not divergences at the surface ; @xmath97 has a maximum at @xmath98 with the positive value @xmath99 , and a minimum at @xmath100 with the negative value @xmath101 .
the width of the curve , that can be estimated as the distance between the two inflection points around the maximum , is @xmath102 . for an increasing cut - off frequency @xmath11
, the curves in figure approach the ideal conductor limit for large distances from the interface , but significant differences still remain close to the surface .
in fact , while the ideal conductor limit @xmath95 diverges with positive values at the surface , @xmath94 assumes more and more negative values as @xmath103 , and the width of the curve reduces to zero .
maximum and minimum values of @xmath94 tend to collapse each other at the surface in the ideal conductor limit , yielding a surface divergence containing a nonvanishing electric and magnetic energy @xmath104 { } z_\eta^{\mathrm{min}}=0.\ ] ] thus , by analyzing the limiting case @xmath15 ( @xmath82 ) it becomes evident that the well - known diverging behavior as @xmath2 near the surface for the ideal conductor , indeed originates from the maximum of @xmath94 .
this is clearly shown in figure [ figcomdivergences ] . the negative divergence at @xmath16 when @xmath82 is completely lost if the calculation is directly performed for an ideal conductor .
our approach thus makes clear the origin of the surface divergences of the renormalized squares of electric and magnetic field ; these divergences , and their detailed structure , naturally appear in the limit process from the dielectric to the ideal conductor that we have used .
the limit @xmath15 is well defined in our model for any distance @xmath3 except @xmath16 . also , due to ,
both integrals over all vacuum space of the ( renormalized ) electric and magnetic energy densities vanish , as expected from the evaluation of global field energies .
this feature does not seem to occur if the calculation is performed directly for an ideal boundary or the plasma model is used even in the limit of an infinite plasma frequency ( see @xcite ) , because in these cases the surface divergences are not fully included in the expressions of the field energy densities . by taking into account the surface divergences with our limit procedure , we thus obtain consistency between global field energies and ( integrated ) local energy densities . field fluctuations and energy densities near the surface can be investigated through the retarded casimir - polder interaction energy with an appropriate polarizable body with static polarizability @xmath105 .
when the distance @xmath106 between the surface and this body is larger that the wavelength associated to its main transition , the so - called far zone approximation holds ; in this case the casimir - polder energy is given by @xmath107 ( see for example @xcite ) .
a similar expression holds for the magnetic fluctuations .
in this paper we have considered zero - point electric and magnetic field fluctuations ( or equivalently field energy densities ) in the vacuum space , when half - space is filled with a homogeneous non - dissipative dielectric .
the material is characterized by a constant real refractive index @xmath7 .
we have then taken the limit @xmath10 and , by introducing a time - splitting procedure with a parameter @xmath62 , we have mathematically included also a high - frequency exponential cut - off function characterized by a cut - off frequency @xmath108 .
we have considered in detail two successive limits : @xmath10 with @xmath14 finite , and then @xmath15 ( ideal conductor limit ) .
we have found that no divergences in the renormalized field fluctuations ( and in the renormalized field energy densities ) are present if @xmath14 is kept finite . in this case ,
field fluctuations at small distances from the interface significantly differ compared to known results for a perfectly conducting plate , as shown in figures [ figdivergences ] and [ figcomdivergences ] . in the limit @xmath15 ( ideal conductor limit ) surface divergences of the field fluctuations ( or energy densities ) at the interface @xmath16 are found , while for @xmath109 the fluctuations approach the well - known @xmath2 behavior of the perfect conductor case .
the structure of the surface divergences has been discussed in detail .
these surface divergences of field fluctuations are not obtained if the field is directly quantized in the presence of an ideal metallic surface ; also the plasma model does not seem to allow to obtain their structure . thus our approach of obtaining the ideal metal through an appropriate limit procedure starting from a dielectric has allowed us to obtain in a natural way the explicit structure of the surface divergence at the interface between vacuum and a conducting material .
having taken into account the field energy in the surface divergences has also allowed us to show consistence between global and local field energies .
finally , we have stressed that field fluctuations and energy densities near the interface can in principle be investigated through the casimir - polder interaction energy with an appropriate electrically or magnetically polarizable body placed near the interface .
the authors acknowledge support from the esf research networking program casimir . financial support by ministero dellistruzione , delluniversit e della ricerca and by comitato regionale di ricerche nucleari e di struttura della materia is also acknowledged .
milonni , _ the quantum vacuum : an introduction to quantum electrodynamics _ , academic press , san diego 1994 .
g. compagno , r. passante , and f. persico , _ atom - field interactions and dressed atoms _ , cambridge university press , cambridge 1995 .
casimir , proc .
akad . wet . *
51 * , 793 ( 1948 ) .
casimir and d. polder , phys . rev .
* 73 * , 360 ( 1948 ) .
r. messina , r. passante , l. rizzuto , s. spagnolo , and r. vasile , j. phys .
a : math . theor . *
164031 ( 2008 ) , and references therein . s. spagnolo , r. passante , and l. rizzuto , phys .
a * 73 * , 062117 ( 2006 ) .
r. passante and s. spagnolo , phys .
a * 76 * , 042112 ( 2007 ) .
milton , i. cavero - pelaez , and j. wagner , j. phys .
a * 39 * , 6543 ( 2006 ) .
milton , in : _ casimir physics _ , edited by d. dalvit , p. milonni , d. roberts , and f. da rosa , springer , heidelberg 2011 , p.39 .
milton , phys .
d * 84 * , 065028 ( 2011 ) a.d .
helfer and a.s.i .
lang , j. phys .
a : math . gen . * 32 * , 1937 ( 1999 ) .
v. sopova and l.h .
ford , phys . rev .
d * 66 * , 045026 ( 2002 ) .
pfenning , phys .
d * 62 * , 045018 ( 2000 ) .
milton , j. phys .
gen * 37 * , 6391 ( 2004 ) . v. sopova and l.h .
ford , phys .
d * 72 * , 033001 ( 2005 ) .
ford and n.f .
svaiter , phys .
d * 58 * , 065007 ( 1998 ) .
mazzitelli , j.p .
nery , a. satz , phys .
d * 84 * , 125008 ( 2011 ) d.j .
alton , n.p .
stern , t. aoki , h. lee , e. ostby , k.j .
vahala , and h.j .
kimble , nature physics * 7 * , 159 ( 2011 ) .
carniglia and l. mandel , phys .
d * 3 * , 280 ( 1971 ) .
i. bialynicki - birula and j.b .
brojan , phys .
d * 5 * , 485 ( 1972 ) .
a.m. contreras reyes and c. eberlein , phys .
a * 80 * , 032901 ( 2009 ) .
a.m. contreras reyes and c. eberlein , phys .
a * 79 * , 043834 ( 2009 ) .
r. passante , e.a .
power , and t. thirunamachandran , phys .
a * 249 * , 77 ( 1998 ) . | we consider the electric and magnetic field fluctuations in the vacuum state in the region external to a half - space filled with a homogeneous non - dissipative dielectric .
we discuss an appropriate limit to an ideal metal and concentrate our interest on the renormalized field fluctuations , or equivalently to renormalized electric and magnetic energy densities , in the proximity of the dielectric - vacuum interface .
we show that surface divergences of field fluctuations arise at the interface in an appropriate ideal conductor limit , and that our limiting procedure allows to discuss in detail their structure .
field fluctuations close to the surface can be investigated through the retarded casimir - polder interaction with an appropriate polarizable body . |
query log analysis has received extensive research attention nowadays , since the exploitation of user feedbacks from query log has been proven to be an effective and non - intrusive method to improve search quality .
search engine has been recording user click through information all the time , which can be represented as a bipartite graph .
the bipartite graph refers to query and url in most cases .
an edge connects a query and a url and the edge value generally corresponds to the click frequency .
many query log analysis models are based on the click graph , in that a certain url has been clicked by different queries ( issued by users ) and hence provides the information about the relevance of url , query and user .
a query can be represented as a vector in which each dimension corresponds to the edge value between the query and a url .
traditional models make use of the raw click frequency ( the number of clicks or users ) between a query and a url , which suffers from two problems : first , raw click frequency does not favor unpopular queries or urls ; moreover , ranking models based on raw click frequency often favor already frequently clicked urls because of inherent bias of clicks @xcite .
therefore it is worth to research on how to improve the representation of click graph before developing any analysis method . to leverage
the influence of highly clicked urls , an entropy - biased model in @xcite has been proposed to address the disadvantages of raw click frequency by weighting the raw click frequency with _ inverse query frequency _ ( iqf ) , under the assumption that less clicked urls are more relevant to a given query than heavily clicked ones .
the inverse query frequency is inspired by _
inverse document frequency _
( idf ) in text retrieval , and @xcite incorporates inverse query frequency into user frequency in the same manner as tf - idf does .
although there are many interpretations @xcite about why tf - idf has proved to be extraordinary robust in text retrieval , utilizing user click frequency in the same manner as tf - idf may not be appropriate in the context of user click graph .
different from content - aware text retrieval , the click through information is the implicit feedback from users in that each click denotes a potential association between a query and a url .
a click tends to be more informative than the case in text mining , since a document could contain much irrelevant information .
therefore , our observation is that user frequency and inverse query frequency should be treated differently during query representation in the context of click graph .
consistent with the assumption that less clicked urls tend to be more relevant to a given query , the inverse query frequency should be more informative than user frequency according to our observation .
moreover , if inverse query frequency can be considered as a global property of each url on click graph , it is intuitive to develop the global consistency model for query representation , which utilizes user frequency and the global weight of url on user click graph in a consistent way to achieve better performance , as described in this paper .
the contribution of this paper lies in : 1 ) we observe that the global nature of the url plays a central role for query representation on user click graph ; 2 ) a new scheme called _ inverse url frequency _ ( iuf ) is presented to specify the global weight of each url on click graph , and result shows that iuf is superior to iqf in the context of global consistency on click graph ; 3 ) we define the rules for achieving global consistency on click graph , and develop the framework of global consistency model for query representation .
the rest of this paper is organized as follows : the related works are introduced in section [ sec : rl ] , and we illustrate global consistency on click graph in section [ sec : gcongraph ] .
various query representation models are presented in section [ sec : queryrepresent ] , while section [ sec : datacollection ] and section [ sec : exp ] presents the experimental analysis about the performance of different models under query similarity .
conclusion is made in section [ sec : conclusion ] .
extensive research has been conducted on click graphs to exploit implicit feedback @xcite .
frequently studied topics include agglomerative clustering @xcite , query clustering for url recommendation @xcite , query suggestion @xcite , which used hitting time to generate semantic consistent suggestions , and rare query suggestion @xcite . moreover , @xcite worked on query classification through increasing the amount of training data by semi - supervised learning on the click graph instead of enriching feature representation
. while there are works studying different aspects of user click information , @xcite revealed that the click probability of a webpage is influenced by its position on the result page .
the sequential nature of user clicks has been considered in @xcite , whereas @xcite combined both the click and skip information from users . in addition , having noticed that click graphs are very sparse and the click frequency follows the power law , @xcite made use of co - click information for document annotation .
random walk has been applied to click graphs @xcite to improve the performance of image retrieval .
@xcite also employed random walk to smooth the click graph to tackle the sparseness issue .
in contrast , less work has been carried out on the study of query representation on click graphs .
@xcite represented each query as a point in a high dimensional space , with each dimension corresponding to a distinct url .
@xcite introduced the query - set based model for document representation using query terms as features for summarizing the clicked webpages .
the entropy - biased model for query representation has been proposed @xcite to replace raw click frequency on the click graph .
it assumed that less clicked urls are more effective in representing a given query than heavily clicked ones .
thus , the raw click frequency was weighed by the inverse query frequency of the url . however , the entropy - biased model utilized raw click frequency and inverse query frequency in the same manner as tf - idf does , which may not be appropriate in the context of click graph .
this is because user click information is content - ignorant while text retrieval is content - aware .
our work is closely related to @xcite , while our contribution is to study how to combine the raw click frequency and the global weight of url in a consistent way for query representation .
the user click graph is generally regarded as a bipartite graph . in this paper
, we consider a bipartite graph evolves query and url : g = ( q @xmath0 d , e ) , where the query set q and document ( url ) set d are connected by edges in e. suppose there are m queries and n documents in total , the bipartite graph can be represented as a rank m@xmath1n matrix c , with the entry ( i , j ) as the edge value of ( @xmath2 , @xmath3 ) . in most cases , the edge value corresponds to the raw click frequency @xmath4 between query @xmath2 and document ( url ) @xmath3 , which is the number of times the users click on @xmath3 when @xmath3 is presented to the users as a result for query @xmath2 .
thus , a query @xmath2 can be represented as a row vector of c , and a document ( url ) @xmath5 corresponds to a column vector of c. .preliminaries and notations [ cols="^,<",options="header " , ] [ tab : length - pagerank ] the average precisions of different models on personalized pagerank are very close compared with the performance on cosine or jaccard similarities , instead it is the behavior of each model that draws our attention . with more steps walked on the click graph
, the result approaches to the stationary distribution .
when the jumping constant @xmath6 , as indicated in figure [ fig : pagerank](a ) , the global consistency model ( ufw - iqf ) experiences a significant drop of average precision at step 2 , while the behavior of other models follows the convention with better precision at step 2 . if the jumping constant @xmath7 which implies an extremely low rate of propagation on the graph , and the first step becomes the dominant factor since the result varies little in subsequent steps .
as shown in figure [ fig : pagerank](b ) , the average precision of each model is determined by the first step .
the personalized pagerank reveals the nature of those different models : for the global consistency model , most of the relevant queries have already been retrieved during the first step , while subsequent steps are needed for other models to retrieve the most relevant results with a reasonable jumping constant . in other words ,
the conventional models need a proper propagation rate to walk on the click graph in order to achieve the best performance , while within global consistency model , the best results are closely clung to the initial preferred vertex without much propagation . in table
[ tab : length - pagerank ] , we also listed the average length of results at rank 10 with personalized pagerank , we observe that on one hand , a smaller jumping constant ( lower propagation rate ) favors long tail results within each model , since the propagation on the graph is heading to the queries with larger transition probability which tend to be the popular short tail queries ; on the other hand , the global consistency model ( ufw - iqf ) demonstrates the superiority of boosting long tail queries over other models despite of different jumping constants .
in this paper we propose a novel model for query representation on user click graph , which is based on the observation that for a certain query , the global nature of urls is more informative than local user frequency .
the global consistency model identifies the inverse query frequency from previous work as a global property of the url , based on which we suggests a more effective scheme called inverse url frequency which further considers the similarities among queries for global nature capturing . besides , we formalize the framework of utilizing user frequency in tune with the global nature of the url .
the global consistency model consistently demonstrates better performance over current models for query representation under popular query similarities , in terms of better precision and long tail result boost .
hence many query log analysis tasks can benefit from this query representation model .
this work is supported by the national natural science foundation of china ( grant nos .
61103185 , 61073118 and 61003247 ) , the start - up foundation of nanjing normal university ( grant no .
2011119xgq0072 ) , natural science foundation of the higher education institutions of jiangsu province , china ( grant no .
11kjb520009 ) , the 9th six talents peak project of jiangsu province ( grant no .
dzxx-043 ) , and major program of national natural science foundation of jiangsu province ( grant no .
bk2011005 ) . | extensive research has been conducted on query log analysis .
a query log is generally represented as a bipartite graph on a query set and a url set .
most of the traditional methods used the raw click frequency to weigh the link between a query and a url on the click graph . in order to address the disadvantages of raw click frequency
, researchers proposed the entropy - biased model , which incorporates raw click frequency with inverse query frequency of the url as the weighting scheme for query representation . in this paper , we observe that the inverse query frequency can be considered a global property of the url on the click graph , which is more informative than raw click frequency , which can be considered a local property of the url . based on this insight
, we develop the global consistency model for query representation , which utilizes the click frequency and the inverse query frequency of a url in a consistent manner .
furthermore , we propose a new scheme called inverse url frequency as an effective way to capture the global property of a url .
experiments have been conducted on the aol search engine log data .
the result shows that our global consistency model achieved better performance than the current models . |
as explained by the presentations of hidalgo and bernal in this conference series , the inclusion of milgrom s acceleration constant @xmath0 as a fundamental quantity in a description of gravity leads to a great deal of understanding in a wide variety of non - relativistic astrophysical systems . as such , let us assume that a point mass @xmath1 located at the origin of coordinates generates a relativistic gravitational field in the mondian regime and that a metric formalism describes the field equations .
this problem is characterised by the speed of light in vacuum @xmath2 , the mass @xmath1 of the central object generating the gravitational field , newton s constant of gravity @xmath3 and milgrom s acceleration constant @xmath4 . with these parameters ,
two `` fundamental lengths '' can be built : @xmath5 where @xmath6 is the gravitational radius and @xmath7 is the mass - length scale as described in @xcite and by hidalgo s presentation in this conference series .
when relativistic effects are taken into account for the gravitational field , then standard general relativity should be recovered in the limit @xmath8 , and a relativistic version of mond should be obtained when @xmath9 .
this shows that the pursue of a complete metric description leads one to consider the scale - dependence of gravity .
the length scales presented in equation must somehow appear in a relativistic theory of gravity which includes the fundamental nature of the constant @xmath4 .
for example , in the metric formalism , a generalised hilbert action @xmath10 can be written in the following way : @xmath11 which slightly differs from its traditional form ( see e.g. @xcite ) since we have introduced the dimensionless ricci s scalar @xmath12 , where @xmath13 is the standard ricci s scalar and @xmath14 defines a length fixed by the parameters of the theory .
the explicit form of the length @xmath14 has to be obtained once a certain known limit of the theory is taken , usually a non - relativistic limit .
in fact , since equation defines two natural lengths of the theory it is coherent to postulate the following relation between all these lengths : @xmath15 note that the definition of @xmath16 gives a correct dimensional character to the action , something that is not completely clear in all previous works dealing with a metric description of the gravitational field . for @xmath17 , the standard einstein - hilbert action is obtained . using the usual form for the matter action @xmath18 , with a matter lagrangian @xmath19 ,
then the null variations of the complete action @xmath20 yield the following field equations : @xmath21 where the dimensionless ricci tensor @xmath22 and @xmath23 is the standard ricci tensor .
the laplace - beltrami operator has been written as @xmath24 and the prime denotes derivative with respect to its argument .
the energy - momentum tensor @xmath25 is defined through the following standard relation : @xmath26 . in here and in what follows
, we choose a ( @xmath27 ) signature for the metric @xmath28 and use einstein s summation convention over repeated indices .
the trace of equation is : @xmath29 where @xmath30 .
under the assumption of a power - law relation for the function @xmath31 , i.e. @xmath32 it follows that the trace can be approximated to order of magnitude since @xmath33 , @xmath34 and the mass density @xmath35 yielding : @xmath36 note that the second term on the left - hand side of equation is much greater than the first term when @xmath37 and so , since @xmath38 , where @xmath39 is the gaussian curvature of space and @xmath40 its radius of curvature , then @xmath41 .
this should occur in the weak - field regime , where mondian effects are expected . for a metric description of gravity
, this limit must correspond to the relativistic regime of mond . in what follows
we will only deal with this approximation and so , equation takes the following form : @xmath42 using the fact that at second order of approximation @xmath43 for a non - relativistic potential @xmath44 and an acceleration @xmath45 , it follows that @xmath46 which converges to a mond - like acceleration regime ( i.e. @xmath47 ) if @xmath48 . at the lowest order of approximation
a description of the gravitational force should not include the velocity of light and so @xmath49 . combining this result with equation
then relation means that @xmath50 which is the traditional form of mond in spherical symmetry . as explained by @xcite in a more rigorous form ,
the result obtained in equation is of general character , since it can be formally proved at the lowest order of perturbation for a @xmath51 metric theory of gravity far away from the mass sources , i.e. when @xmath52 , or equivalently when @xmath53 . in @xcite
it is also discussed the noether symmetries of the problem and for this particular case the conserved charge of the problem is proportional to @xmath54 .
the metric theory of gravity presented here is by no means a complete description at all scales of gravitation .
it only deals with the mondian regime of gravity , i.e. when @xmath53 is valid . in other words ,
our description breaks the scale invariance of gravity in a more general way than the one described in @xcite .
the mass dependence of @xmath55 means that the mass needs to appear on hilbert s action .
this is traditionally not the case , since that action is thought to be purely a function of the geometry of space - time due to the presence of mass and energy .
however , it was @xcite who first encountered this peculiarity in the hilbert action when dealing with a metric generalisation of mond and later on also discussed by @xcite . following the remarks of @xcite
, one should not be surprised if some of the commonly accepted notions , even at the fundamental level of the action , require generalisations and re - thinking .
an extended metric theory of gravity goes beyond the traditional general relativity ideas and in this way , we probably need to change our standard view of its fundamental principles . with these ideas in mind
it is interesting to note that recently @xcite have proposed a modified @xmath56 theory of gravity , where @xmath57 represents the trace of the energy - momentum tensor .
the theory developed in this article is a particular example of their proposal with the identification @xmath58 .
we are developing a cosmological theory with these ideas in mind and we are also applying this full theory to gravitational lenses and to the dynamics of clusters of galaxies .
the obtained results will be reported elsewhere .
this work was supported by a dgapa - unam grant ( papiit in116210 - 3 ) .
the authors tb , jch & sm acknowledge economic support from conacyt : 207529 , 51009 , 26344 .
40 s. mendoza , x. hernandez , j. c. hidalgo and t. bernal , mon . not .
* 411 * ( 2011 ) 226 s. capozziello and v. faraoni , beyond einstein gravity : a survey of gravitational theories for cosmology and astrophysics , springer,(2010 ) isbn 9789400701649 t. bernal , s. capozziello , j.c .
hidalgo and s. mendoza , european physical journal c * 71 * ( 2011 ) 1774 y. sobouti , astronomy & astrophysics , * 464 * ( 2007 ) 921 s. mendoza and y. m.rosas-guevara , astronomy & astrophysics ( 2007 ) * 472*,367 t. harko , f. s.n .
lobo , s. nojiri and s. d. odintsov , phys.rev.d ( 2011 ) 84:024020 | we show that the modified newtonian dynamics ( mond ) regime can be fully recovered as the weak - field limit of a particular theory of gravity formulated in the metric approach .
this is possible when milgrom s acceleration constant is taken as a fundamental quantity which couples to the theory in a very consistent manner . as a consequence , the scale invariance of the gravitational interaction is naturally broken . in this sense ,
newtonian gravity is the weak - field limit of general relativity and mond is the weak - field limit of that particular extended theory of gravity .
address = instituto de astronoma , universidad nacional autnoma de mxico , ap 70 - 264 , distrito federal 04510 , mxico .
address = instituto de astronoma , universidad nacional autnoma de mxico , ap 70 - 264 , distrito federal 04510 , mxico .
address = instituto de astronoma , universidad nacional autnoma de mxico , ap 70 - 264 , distrito federal 04510 , mxico .
address = dipartimento di scienze fisiche , universit degli studi di napoli `` federico ii '' , complesso universitario di monte santangelo , edificio n , via cinthia , 80126 , napoli , italy + i.n.f.n .
- sezione di napoli , complesso universitario di monte santangelo , edificio g , via cinthia , 80126 , napoli , italy |
the goal of the present paper is to show that : @xmath0 eilenberg s theorem is an important result in algebraic language theory , stating that there is a one to
one correspondence between certain classes of regular languages , called varieties of languages , and certain classes of monoids , called pseudovarieties of monoids ( * ? ? ?
* theorem 34 ) .
the concept of regular language , which is defined in terms of deterministic automata , has an equivalent machine independent algebraic definition , namely , a language recognized by a finite monoid .
recognizable languages on an alphabet @xmath5 are inverse images of monoid homomorphisms with domain @xmath6 and as codomain any finite monoid .
this algebraic approach allows us to study various kinds of recognizable languages where the notion of homomorphism between algebras is a key ingredient .
the study of algebras and classes of algebras is a main subject of study in universal algebra . a well known theorem in this area is birkhoff s variety theorem @xcite , which states that a class of algebras of a given type is defined by a set of equations if and
only if it is a variety , i.e. , it is closed under homomorphic images , subalgebras , and products .
later , a birkhoff s theorem for finite algebras was also obtained @xcite , also known as reiterman s theorem . in birkhoff s theorem for finite algebras the kind of equations are of a more general kind , they can be defined by using topological techniques or can be equivalently defined by using the so called implicit operations . in this case , the classes of algebras considered are pseudovarieties of algebras , also known as varieties of finite algebras , which are defined as classes of finite algebras of the same type that are closed under homomorphic images , subalgebras and finite products . to state eilenberg
type theorems , which establish one to one correspondences between ( pseudo)varieties of algebras and ( pseudo)varieties of languages , one has to define and find the corresponding notion of a ( pseudo)variety of languages which is , in general , a non trivial problem .
there are eilenberg type correspondences in the literature such as , e.g. , @xcite for pseudovarieties of ordered monoids and ordered semigroups , the one in @xcite for pseudovarieties of finite dimensional @xmath7algebras , @xcite for pseudovarieties of idempotent semirings and ( * ? ? ?
* theorem 39 ) for varieties of monoids .
the work in the present paper has its basis in @xcite .
we take the main idea given in @xcite , where algebras for a monad @xmath1 on @xmath2 are considered , to define the natural notion of a ( pseudo)variety of @xmath1algebras . in order to characterize the kind of equations defining a ( pseudo)variety of @xmath1algebras
, we use the natural approach of capturing equations as epimorphisms in @xmath2 , i.e. , congruences , and add the condition that those equations are closed under substitution .
these properties are captured in categorical terms to define the notion of a ( pseudo)equational @xmath1theory .
we obtain that , under mild assumptions , ( pseudo)varieties of @xmath1algebras are exactly classes of ( finite ) @xmath1algebras that are defined by ( pseudo)equational @xmath1theories , and that they are in one to
one correspondence .
this will give us a categorical version of birkhoff s theorem for ( finite ) @xmath1algebras .
once we get this one to one correspondence between ( pseudo)varieties of @xmath1algebras and ( pseudo)equational @xmath1theories , we use a category @xmath3 that is dual to @xmath2 and a result given in @xcite that allows us to define a canonical comonad @xmath4 on @xmath3 that is dual to @xmath1 and lift the duality between @xmath3 and @xmath2 to their corresponding eilenberg moore categories . with this duality
, there is a canonical correspondence between ( pseudo)equational @xmath1theories and their corresponding dual , i.e. , ( pseudo)coequational @xmath4theories .
our most important examples of ( pseudo)coequational @xmath4theories are those given in eilenberg type correspondences , i.e. , `` varieties of languages '' .
all in all , we get a one to one correspondence between ( pseudo)varieties of @xmath1algebras and ( pseudo)coequational @xmath4theories .
we will show how this concept of ( pseudo)coequational @xmath4theories coincides with the different notions of `` varieties of languages '' in eilenberg type correspondences , which bring us to our slogan , eilenberg type correspondences = birkhoff s theorem for ( finite ) algebras + duality . as a consequence , we can summarize eilenberg
type correspondences in the following picture : ( ) at ( 0,-0.05 ) ( pseudo)varieties ; ( ) at ( 0,-0.45 ) of @xmath1algebras ; ( 1.5,-0.25 ) ( 7.5,-0.25 ) ; ( ) at ( 4.5,0 ) eilenberg type correspondences ; ( ) at ( 11.6,0.3 ) @xmath8 ; ( ) at ( 9.5,-0.2 ) @xmath9 ; where ` @xmath8 ' denotes the dual operator .
this easy to understand and straightforward one to one correspondence gives us what we called an abstract eilenberg type correspondence for ( pseudo)varieties of @xmath1algebras , proposition [ eilvar ] and [ eilpvar ] , from which we recover and discover particular instances of eilenberg type correspondences for different kinds of algebraic structures , i.e. , @xmath1algebras .
it is worth mentioning that eilenberg
type correspondences have not been fully understood for the last forty years , which can be witnessed by the numerous published results on the subject that deal with specific kinds of algebras such as @xcite and categorical generalizations such as @xcite in which the direct relation between `` varieties of languages '' and equational theories , by using duality , is not studied or explored to find and justify the defining properties of a `` variety of languages '' . * related work .
* we briefly summarize here some related work ( see the conclusions for a more detailed discussion . )
there are various generalizations of birkhoff s theorem for ( finite ) algebras such as @xcite . in order to derive eilenberg
type correspondences , in this paper we prove a categorical versions of birkhoff s theorem for algebras and finite algebras , which are stated , under mild assumptions , as one to one correspondences between ( pseudo)varieties of algebras and ( pseudo)equational theories .
the variety version is derived from @xcite and the pseudovariety version is based on the observation that pseudovarieties of algebras are directed unions of of equational classes of finite algebras ( * ? ? ?
* proposition 4 ) .
it is worth mentioning that the proof presented here for birkhoff s theorem for finite @xmath1algebras does not involve the use of topology nor profinite techniques , in contrast to @xcite .
related work such as @xcite have influenced and motivated the use of duality in language theory to characterize recognizable languages and to derive local versions of eilenberg type correspondences . in the present paper ,
the use of duality is a key aspect that helps us to understand and unveil eilenberg
type correspondences .
there are some works in which categorical approaches to derive eilenberg
type correspondences are used , notably @xcite .
the work in @xcite subsumes the work made in @xcite and the present paper subsumes the work made in @xcite .
the kind of algebras considered in @xcite are algebras with a monoid structure which restricts the kind of algebras one can consider , e.g. , eilenberg s theorem ( * ? ? ?
* theorem 34s ) for pseudovarieties of semigroups can not be derived from @xcite . a different approach to get a general eilenberg
type theorem is the approach given in @xcite were the algebras considered are algebras for a monad @xmath1 on @xmath10 , for a fixed set @xmath11 . the fact that all the monads considered are on @xmath10 was not general enough to cover cases such as @xcite in which the varieties of languages are not necessarily boolean algebras .
the approach in @xcite of considering algebras for a monad @xmath1 is also considered and generalized in @xcite as well as in the present paper .
one of the main challenges in categorical approaches to eilenberg type correspondences is to define the right concept of a `` variety of languages '' .
the definition of a `` variety of languages '' given in @xcite depends of finding what they call a `` unary representation '' , which is a set of unary operations on a free algebra satisfying certain properties , see ( * ? ? ?
* definition 3.7 . ) . from this `` unary representation ''
one can construct syntactic algebras and define the kind derivatives that define a `` variety of languages '' .
the definition of a `` variety of languages '' in the present paper is a categorical one which avoids the explicit definition of derivatives and existence of syntactic algebras . in the present paper , derivatives
are captured coalgebraically and syntactic algebras are not used to prove the abstract eilenberg type correspondences theorems , but both of those concepts can be easily obtained via duality in each concrete case .
coalgebraic approaches , from which one can easily define the concept of a `` variety of languages '' , are not used in @xcite .
another important related work is @xcite , in which an eilenberg type correspondence for varieties of monoids is shown , which is an eilenberg type correspondence that can be derived from the present paper but not from @xcite . the work made in @xcite motivates the study of eilenberg type correspondences for other classes of algebras different than pseudovarieties .
it is worth mentioning that in @xcite the duality between equations and coequations is studied for the first time in the context of an eilenberg type correspondence .
all in all , the contributions of the present paper can be summarized as follows : 1 .
to unveil eilenberg
type correspondences and show that : @xmath0 2 . to show and understand where `` varieties of languages '' come from ,
that is : @xmath12 this fact was conjectured by the author in @xcite .
3 . to provide categorical versions of birkhoff s theorem for ( finite ) @xmath1algebras as one to
one correspondences between ( pseudo)varieties of @xmath1algebras and ( pseudo)equational @xmath1theories , which are easily obtained from @xcite . a categorical definition of a ( pseudo)equational theory is given .
4 . to show a categorical version of birkhoff s theorem for finite algebras without the use of topology nor profinite techniques .
5 . to provide a general and abstract eilenberg type correspondence theorem that encompasses existing eilenberg type correspondences from the literature . not only for ( local ) pseudovarieties of algebras but also for ( local ) varieties of algebras . 6 . to derive eilenberg
type correspondences without the use of syntactic algebras . 7 . to show that the notion of derivatives used to define the different kinds of `` varieties of languages '' in eilenberg type correspondences is exactly the coalgebraic structure of an object , which is easily derived via duality , is most of the cases , from the notion of an algebra homomorphism .
we present the content of our paper as follows : in section 1 , we fix some notation and some categorical facts we will use through the paper . in section 2 ,
we state an abstract eilenberg type correspondence for varieties of @xmath1algebras , which is derived from a categorical version of birkhoff s theorem plus duality . as an application ,
we derive eilenberg type correspondences for some varieties of algebras . in section 3
, we redo section 2 , for the case of pseudovarieties of @xmath1algebras , i.e. , we use birkhoff s theorem for finite @xmath1algebras . in section 4 , we do a similar work for local ( pseudo)varieties of @xmath13generated @xmath1algebras whose proof easily follows from what is done in section 2 and section 3 .
then we finish in section 5 with the conclusions of this work .
we introduce the notation and some facts that we will use in the paper .
we assume that the reader is familiar with basic concepts from category theory and ( co)algebra , see , e.g. , @xcite . given a category @xmath2 and a monad @xmath14 on @xmath2
, we denote the category of ( eilenberg moore ) @xmath1-algebras and their homomorphisms by @xmath15 .
objects in @xmath15 are pairs @xmath16 where @xmath13 is an object in @xmath2 and @xmath17 is a morphism @xmath18 in @xmath2 that satisfies the identities @xmath19 and @xmath20 . a homomorphism from a @xmath1-algebra @xmath21 to a @xmath1-algebra @xmath22 is a morphism @xmath23 such that @xmath24 .
dually , given a category @xmath3 and a comonad @xmath25 on @xmath3 , @xmath26 denotes the category of ( eilenberg moore ) @xmath4-coalgebras .
objects in @xmath26 are pairs @xmath27 where @xmath28 is an object in @xmath3 and @xmath29 satisfies the identities @xmath30 and @xmath31 .
a homomorphism from a @xmath4-coalgebra @xmath32 to a @xmath4-coalgebra @xmath33 is a morphism @xmath34 such that @xmath35 .
if @xmath2 and @xmath3 are dual categories and @xmath1 is a monad on @xmath2 , then there is a canonical comonad @xmath4 on @xmath3 such that the duality between @xmath2 and @xmath3 lifts to their corresponding eilenberg moore categories .
[ monadtocomonad ] ( * ? ? ?
* proposition 14 ) let @xmath36 and @xmath37 be contravariant functors that form a duality with natural isomorphisms @xmath38 and @xmath39 .
let @xmath40 be a monad on @xmath2 .
then @xmath25 , where @xmath41 and @xmath42 are defined as : @xmath43 is a comonad on @xmath3 .
further , the duality between @xmath44 and @xmath45 lifts to a duality between @xmath46 and @xmath47 .
the following is a list of the categories we will use in the examples given in this paper : @xmath48 & sets & functions + @xmath49 & complete atomic boolean algebras & complete boolean algebra homomorphisms + @xmath50 & partially ordered sets & order preserving functions + @xmath51 & algebraic completely distributive lattices & complete lattice homomorphisms + @xmath52 & @xmath7vector spaces & linear maps + @xmath53 & topological @xmath7vector spaces that are stone spaces , i.e , they have compact , hausdorff and zero dimensional topology & linear continuous maps + @xmath54 & join semilattices with zero & join semilattice homomorphisms preserving zero + @xmath55 & topological join semilattices with zero that are stone spaces & continuous join semilattice homomorphisms preserving zero + we will use the facts that @xmath48 is dual to @xmath49 , @xmath50 is dual to @xmath51 , @xmath52 is dual to @xmath53 for a finite field @xmath7 and that @xmath54 is dual to @xmath55 . for a given concrete category @xmath3
, we denote the full subcategory of @xmath3 consisting of its finite objects by @xmath56 . in the case of @xmath15 and @xmath26 , we denote them by @xmath57 and @xmath58 , respectively .
let @xmath2 be a category and let @xmath59 and @xmath60 be classes of morphisms in @xmath2 .
@xmath61 is called a _ factorization system _ on @xmath2 if : a. each of @xmath59 and @xmath60 is closed under composition with isomorphisms , b. every morphism @xmath62 in @xmath2 has a factorization @xmath63 , with @xmath64 and @xmath65 . c. given any commutative diagram + \(a ) at ( 0,0 ) @xmath66 ; ( b ) at ( 2,0 ) @xmath66 ; ( c ) at ( 0,1.5 ) @xmath66 ; ( d ) at ( 2,1.5 ) @xmath66 ; ( a ) edge [ ] node [ below ] @xmath67 ( b ) ; ( c ) edge [ ] node [ above ] @xmath68 ( d ) ; ( c ) edge [ ] node [ left ] @xmath62 ( a ) ( d ) edge [ ] node [ right ] @xmath69 ( b ) ; + with @xmath64 and @xmath65 , there is a unique _ diagonal fill in _ , i.e. , a unique morphism @xmath70 such that the following diagram commutes : + \(a ) at ( 0,0 ) @xmath66 ; ( b ) at ( 2,0 ) @xmath66 ; ( c ) at ( 0,1.5 ) @xmath66 ; ( d ) at ( 2,1.5 ) @xmath66 ; ( a ) edge [ ] node [ below ] @xmath67 ( b ) ; ( c ) edge [ ] node [ above ] @xmath68 ( d ) ; ( c ) edge [ ] node [ left ] @xmath62 ( a ) ( d ) edge [ ] node [ above ] @xmath70 ( a ) ( d ) edge [ ] node [ right ] @xmath69 ( b ) ; a factorization system @xmath61 is _ proper _ if every morphism in @xmath59 is epi and every morphism in @xmath60 is mono .
we will use the following facts about factorization systems @xcite .
[ lemmafs ] let @xmath2 be a category and @xmath61 be a factorization system on @xmath2 such that every morphism in @xmath60 is mono
. then @xmath71 implies @xmath72 .
[ lemmafs1 ] let @xmath2 be a category , @xmath14 a monad on @xmath2 and @xmath61 a proper factorization system on @xmath2 .
if @xmath73 preserves the morphisms in @xmath59 then @xmath15 inherits the same @xmath61 factorization system .
varieties of algebras have been studied in universal algebra and equational logic . in particular , birkhoff s variety theorem ( see , e.g. , @xcite ) states that a class of algebras of the same type is a variety , i.e. , it is closed under homomorphic images , subalgebras and ( not necessarily finite ) products , if and only if it is definable by equations . as a consequence , for a fixed type of algebras , we get a one to one correspondence between varieties of algebras and equational theories .
birkhoff s theorem has been generalized to a categorical level , see , e.g. , @xcite , to characterize subcategories of a given category that are , in some sense , equationally defined . in this section ,
we provide , under mild assumptions , a birkhoff s theorem for varieties of @xmath1algebras , theorem [ birkthm ] , which , in order to derive eilenberg type correspondences , will be stated as a one to one correspondence between varieties of @xmath1algebras and equational @xmath1theories . a categorical definition of an equational @xmath1theory will be given .
next we dualize the categorical definition of an equational @xmath1theory to get that of a coequational @xmath4theory , where @xmath4 is a comonad on a category @xmath3 .
our main contribution is to note that particular instances of coequational @xmath4theories have been already studied in the literature under the name of `` varieties of languages '' to establish eilenberg type correspondences , e.g. , ( * ? ? ?
* theorem 39 ) .
thus , if we assume that @xmath2 and @xmath3 are dual categories and that the comonad @xmath4 is the dual of the monad @xmath1 , as in proposition [ monadtocomonad ] , then , by duality , we get a one to one correspondence between equational @xmath1theories and coequational @xmath4theories .
all in all , we get a one to one correspondence between varieties of @xmath1algebras and coequational @xmath4theories , which is the abstract eilenberg type correspondence for varieties of @xmath1algebras , proposition [ eilvar ] .
the main facts in this section can be summarized in the following picture : ( ) at ( 0,-0.05 ) varieties of ; ( ) at ( 0,-0.45 ) @xmath1algebras ; ( 1,-0.25 ) ( 6,-0.25 ) ; ( ) at ( 3.5,-0.5 ) birkhoff s thm . ;
( ) at ( 3.5,-0.9 ) theorem [ birkthm ] ; ( ) at ( 7,-0.05 ) equational ; ( ) at ( 7,-0.45 ) @xmath1theories ; ( 7,-0.75 ) ( 7,-1.7 ) ; ( ) at ( 7.7,-1.2 ) duality ; ( ) at ( 7,-1.75 ) coequational ; ( ) at ( 7,-2.15 ) @xmath4theories ; ( 0,-0.75 ) ( 0,-1.95)(6,-1.95 ) ; ( ) at ( 2.5,-2.2 ) eilenberg type correspondence ; ( ) at ( 2.5,-2.6 ) proposition [ eilvar ] ; where each arrow symbolizes a one to
one correspondence and @xmath4 is the comonad that is the dual of the monad @xmath1 . the concept of a variety of algebras and an equationally defined class can be formulated in categorical terms to prove a categorical version of birkhoff s theorem @xcite .
we prove in this subsection that , under mild assumptions , there is a one to
one correspondence between varieties of @xmath1algebras and equational @xmath1theories , theorem [ birkthm ] , where @xmath1 is a monad on a category @xmath2 .
the definition of variety of @xmath1algebras , which depends on the concept of homomorphic images and subalgebras , will be defined by using a factorization system @xmath61 on @xmath2 . in order to define the concept of equational @xmath1theories , we base our approach on ( * ? ? ?
* definition ii.14.16 ) . after providing the assumptions and basic definitions needed to state theorem [ birkthm ]
, its proof will easily follow from the assumptions needed by using the work by banaschewski and herrlich @xcite .
we fix a complete category @xmath2 , a monad @xmath14 on @xmath2 , a factorization system @xmath61 on @xmath2 and a full subcategory @xmath74 of @xmath2 .
we will use the following assumptions : 1 .
the factorization system @xmath61 is proper .
that is , every map in @xmath59 is an epimorphism and every map in @xmath60 is a monomorphism .
2 . for every @xmath75
, the free @xmath1algebra @xmath76 is _ projective with respect to @xmath59 in @xmath15_. that is , for every @xmath77 with @xmath75 and @xmath78 there exists @xmath79 such that the following diagram commutes : + \(a ) at ( 0,0 ) @xmath80 ; ( b ) at ( 3,0 ) @xmath81 ; ( c ) at ( 3,1.5 ) @xmath82 ; ( a ) edge [ ] node [ below ] @xmath68 ( b ) ; ( a ) edge [ ] node [ above ] @xmath83 ( c ) ; ( b ) edge [ ] node [ right ] @xmath69 ( c ) ; 3 . for every @xmath84 there exists @xmath85 and @xmath86 .
@xmath73 preserves morphisms in @xmath59 .
5 . for every @xmath75
, there is , up to isomorphism , only a set of @xmath1algebra morphisms in @xmath59 with domain @xmath87 .
the notion of a variety of @xmath1algebras , which depends on the concept of homomorphic images and subalgebras , will be defined by using the factorization system @xmath61 on @xmath2 , which is lifted to @xmath15 using ( b1 ) and ( b4 ) , lemma [ lemmafs1 ] .
the role of @xmath74 is that the objects from which `` variables '' for the equations are considered are objects in @xmath74 .
assumption ( b2 ) of @xmath87 being projective with respect to @xmath59 , @xmath75 , will play a fundamental role in relating varieties of algebras with equational theories . assumption ( b3 )
guarantees that every algebra in @xmath15 is the homomorphic image of a free @xmath1algebra with object of generators from @xmath74 .
condition ( b5 ) will allow us to define the equational theory for a given variety of algebras . for birkhoff s classical variety theorem @xcite , we can take @xmath88 , @xmath89 surjections , @xmath90 injections , and @xmath1 to be the term monad for a given type of algebras @xmath91 , i.e. , @xmath92 , the set of terms of type @xmath91 on the set of variables @xmath13 ( see example [ excleqth ] and example [ exbirkhoff ] )
. another important example will be given by @xmath93 , with @xmath94 discrete posets ( i.e. , we do not want the `` variables '' to be ordered ) to obtain a birkhoff theorem for ordered algebras @xcite .
we now give the necessary definitions to formulate our birkhoff theorem for @xmath1algebras .
we start by defining varieties of @xmath1algebras .
let @xmath2 be a complete category , @xmath1 a monad on @xmath2 and @xmath61 a factorization system on @xmath2 .
let @xmath95 be a class of algebras in @xmath15 .
we say that @xmath95 _ is closed under @xmath59quotients _ if @xmath96 for every @xmath78 with @xmath97 .
we say that @xmath95 _ is closed under @xmath60subalgebras _ if @xmath96 for every @xmath98 with @xmath97 .
we say that @xmath95 _ is closed under products _ if @xmath99 for every set @xmath100 such that @xmath101 , @xmath102 .
a class @xmath103 of algebras in @xmath15 is called a _ variety of @xmath1algebras _ if it is closed under @xmath59quotients , @xmath60subalgebras and products . now , we define the other main concept to state our bikhoff s theorem for @xmath1algebras , namely , the concept of an equational @xmath1theory . [ defeqth ]
let @xmath2 be a category , @xmath1 a monad on @xmath2 , @xmath74 a full subcategory of @xmath2 and @xmath61 a factorization system on @xmath2 .
equational @xmath1theory on @xmath74 _ is a family of @xmath1algebra morphisms @xmath104 in @xmath59 such that for any @xmath105 and any @xmath106 there exists @xmath107 such that the following diagram commutes : \(a ) at ( 2,1.5 ) @xmath108 ; ( b ) at ( 2,0 ) @xmath109 ; ( c ) at ( 0,1.5 ) @xmath82 ; ( d ) at ( 0,0 ) @xmath110 ; ( a ) edge [ ] node [ right ] @xmath111 ( b ) ( c ) edge [ ] node [ left ] @xmath112 ( d ) ; ( c ) edge [ ] node [ above ] @xmath113 ( a ) ; ( d ) edge [ ] node [ below ] @xmath114 ( b ) ; intuitively , in the setting of birkhoff s classical variety theorem , for every object @xmath75 ( i.e. , a set of variables ) the morphism @xmath112 , which we asssume to be a surjection , represents the set of equations @xmath115 , which is a congruence on @xmath87 , i.e. , it is an equivalence relation on @xmath82 which is closed under the componentwise algebric operations .
commutativity of the diagram above means that the family of all equations @xmath116 is closed under any substitution @xmath106 .
the previous definition generalizes the definition of an equational theory to a categorical level , cf .
* definition ii.14.16 ) .
[ excleqth ] consider the case @xmath88 , @xmath89 surjections and @xmath90 injections .
for a given type of algebras @xmath91 , consider the monad @xmath117 such that @xmath118 is the set of terms for @xmath91 on variables @xmath13 , see ( * ? ? ?
* definition ii.10.1 ) .
the unit @xmath119 is the inclusion function and multiplication @xmath120 is the identity map .
now , @xmath121 is the category of algebras @xmath122 of type @xmath91 , where @xmath123 is the evaluation @xmath124 in @xmath81 of each term @xmath125 .
an equational @xmath126theory on @xmath127 is a family of surjective homomorphisms @xmath128 in @xmath121 such that every @xmath129 is a congruence on @xmath118 and the family @xmath116 is closed under substitution , i.e. , for @xmath130 and @xmath131 , @xmath132 , we have that @xmath133 , where @xmath134 is the term in @xmath135 obtained from @xmath136 by replacing each variable @xmath137 by @xmath138 , @xmath139 .
[ exineqth ] consider the case @xmath93 , @xmath94 the full subcategory of discrete posets , @xmath89 surjections and @xmath90 embeddings .
let @xmath91 be a type of algebras
. an _ ordered algebra of type @xmath91 _ is a triple @xmath140 such that @xmath141 and all the functions @xmath142 are order preserving , where the order in @xmath143 is componentwise , @xmath144 .
we can define the monad @xmath117 where @xmath145 is the poset @xmath146 defined as : @xmath147 for every @xmath148 such that @xmath149 , and @xmath150 for every @xmath144 and terms @xmath151 such that @xmath152 , @xmath153 .
algebras in @xmath121 are ordered algebras of type @xmath91 .
an equational @xmath126theory is a family @xmath154 of surjective homomorphisms , which are trivially order preserving since @xmath118 is discrete for any @xmath75 , such that @xmath155 is an admissible preorder on @xmath118 , on an ordered algebra @xmath156 of type @xmath91 is _ compatible _ if for every @xmath144 and @xmath157 with @xmath158 , @xmath159 , we have that @xmath160 . a preorder @xmath161 is _ admissible _ if it is compatible and @xmath162 whenever @xmath163 .
the congruence @xmath164 on @xmath81 induced by the compatible preorder @xmath161 is the relation @xmath164 on @xmath81 defined as @xmath165 .
then @xmath166 is an ordered algebra with the order given by @xmath167\leq_\sqsubseteq [ y]$ ] iff @xmath168 .
see @xcite . ]
where @xmath169 , and the family @xmath170 is closed under substitution as in the previous example . in this case , @xmath155 represents the equations and inequations of terms with variables in @xmath13 in the equational @xmath126theory .
note that if we take @xmath171 then condition ( b2 ) does not hold .
given an equational @xmath1theory @xmath104 and an algebra @xmath84 , we say that @xmath172 _ satisfies _ @xmath173 , denoted as @xmath174 , if @xmath172 is _ @xmath173injective _ , that is , if for every @xmath75 and every @xmath175 there exists a @xmath1algebra morphism @xmath176 such that @xmath177 .
intuitively , @xmath174 if for every assignment @xmath178 of the variables @xmath13 to elements of the algebra @xmath172 , all the equations represented by @xmath179 hold in @xmath172 .
given an equational @xmath1theory @xmath173 we denote the _ models of _
@xmath173 by @xmath180 , that is : @xmath181 a class @xmath95 of @xmath1algebras is _ defined _ by @xmath173 if @xmath182 .
[ birkthm ] let @xmath2 be a complete category , @xmath1 a monad on @xmath2 , @xmath74 a full subcategory of @xmath2 and @xmath61 a factorization system on @xmath2 .
assume ( b1 ) to ( b5 ) .
then a class @xmath95 of @xmath1algebras is a variety of @xmath1algebras if and only if it is defined by an equational @xmath1theory on @xmath74 .
additionally , varieties of @xmath1algebras are in one to
one correspondence with equational @xmath1theories on @xmath74 . from the previous theorem
we have the following . [ exbirkhoff ] by considering the monad and the categories given in example [ excleqth ]
we obtain the classical birkhoff variety theorem @xcite .
[ exorderedbirkhoff ] by considering the monad and the categories given in example [ exineqth ] we obtain the birkhoff variety theorem for ordered algebras @xcite . [ exjan1 ] consider the case @xmath88 , @xmath1 the monad given by @xmath183 , where @xmath184 is the free monoid on @xmath13 , @xmath89 surjections and @xmath90 injections .
we have that conditions ( b1 ) to ( b5 ) are fullfilled
. therefore we have a one to one correspondence between varieties of monoids and equational @xmath1theories .
now , consider the category @xmath185
which is dual to @xmath48 and let @xmath4 be the comonad on @xmath49 that is dual to the monad @xmath1 on @xmath48 , i.e , @xmath80 is defined , up to isomorphism , as @xmath186 . then , by duality , we have a one to one correspondence between equational @xmath1theories @xmath173 and its dual @xmath187 , i.e. , families of monomorphisms @xmath188 in @xmath26 such that for any @xmath189 and any @xmath190 there exists @xmath191 such that @xmath192 .
this notion of the dual of an equational @xmath1theory is equivalent with the more complicated notion of a variety of languages in ( * ? ? ?
* definition 35 ) ( see example [ exjan11 ] for more details ) . in this setting , we get a one to one correspondence between varieties of monoids and duals of equational @xmath1theories , i.e. , varieties of languages .
this is exactly the eilenberg type theorem ( * ? ? ?
* theorem 39 ) . as we saw in example [ exjan1 ] , by using theorem [ birkthm ] and duality , we can derive eilenberg type correspondences for varieties of @xmath1algebras . in this subsection we state the abstract eilenberg type correspondence for varieties of @xmath1algebras , i.e. , a one to one correspondence between varieties of @xmath1algebras and duals of equational @xmath1theories , and then instantiate this result to some particular cases .
we dualize the definition of an equational @xmath1theory as follows .
[ defcoeqth ] let @xmath3 be a category , @xmath25 a comonad on @xmath3 , @xmath61 a factorization system on @xmath3 and @xmath193 a full subcategory of @xmath3 .
coequational @xmath4theory on @xmath193 _ is a family of @xmath4coalgebra morphisms @xmath194 in @xmath60 such that for any @xmath189 and any @xmath190 there exists @xmath191 such that the following diagram commutes : \(a ) at ( 2,0 ) @xmath195 ; ( b ) at ( 2,1.5 ) @xmath196 ; ( c ) at ( 0,0 ) @xmath197 ; ( d ) at ( 0,1.5 ) @xmath198 ; ( a ) edge [ ] node [ right ] @xmath199 ( b ) ( c ) edge [ ] node [ left ] @xmath200 ( d ) ; ( c ) edge [ ] node [ below ] @xmath114 ( a ) ; ( d ) edge [ ] node [ above ] @xmath113 ( b ) ; intuitively , every @xmath201 is a @xmath4subcoalgebra of the cofree coalgebra @xmath202 , and the family @xmath203 is closed under any coalgebra morphism , i.e. , for every morphism @xmath190 , @xmath204 implies @xmath205 .
as an example of coequational @xmath4theories we have the `` varieties of languages '' defined in ( * ? ? ?
* definition 35 ) which we describe in a simpler way in example [ exjan11 ] .
now , with the previous definition , theorem [ birkthm ] and duality , we have the following . [ eilvar ]
let @xmath2 be a complete category , @xmath1 a monad on @xmath2 , @xmath61 a factorization system on @xmath2 and @xmath74 a full subcategory of @xmath2 .
assume ( b1 ) to ( b5 ) .
let @xmath3 be a category that is dual to @xmath2 , @xmath193 the corresponding dual category of @xmath74 and let @xmath4 be the comonad on @xmath3 that is dual to @xmath1 which is defined as in proposition [ monadtocomonad ] .
then there is a one to
one correspondence between varieties of @xmath1algebras and coequational @xmath4theories on @xmath193 . in the rest of this section
we show some particular instances of eilenberg type correspondences derived from the the previous proposition .
we start with the continuation of example [ exjan1 ] by describing the defining properties of the coequational @xmath4theories that correspond to varieties of monoids .
then we do a similar work to derive eilenberg
type correspondences for other kind of varieties such as semigroups , groups , monoid actions , ordered monoids , vector spaces and idempotent semirings .
it is worth mentioning that these kind of results , except for the one in example [ exjan11 ] which was shown in ( * ? ? ?
* theorem 39 ) , seem to be new and can not be derived with categorical approaches given in @xcite . [ exjan11 ] from the setting of example [ exjan1 ] we get a one to one correspondence between varieties of monoids and coequational @xmath4theories on @xmath49 .
the latter can be characterized as operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath208 and it is a subalgebra of the complete atomic boolean algebra @xmath209 of subsets of @xmath184 , i.e. , every element in @xmath210 is a language on @xmath13 .
b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath132 then @xmath212 , where @xmath213 and @xmath214 , @xmath215 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of monoids @xmath217 and @xmath211 , we have that @xmath218 . the previous notion of a coequational @xmath4theory on @xmath49
is equivalent with the more complicated notion of a variety of languages in ( * ? ? ?
* definition 35 ) .
this one to one correspondence is exactly the eilenberg type theorem ( * ? ? ?
* theorem 39 ) ( see example [ exua ] and appendix for more details ) .
the following example describes explicitely eilenberg type correspondences for varieties of algebras of any given type @xmath91 where each function symbol in @xmath91 has finite arity .
[ exua ] let @xmath91 be a type of algebras where each function symbol @xmath219 has arity @xmath220 and let @xmath95 be a variety of algebras of type @xmath91 . consider the case @xmath88 , @xmath89 surjections , @xmath90 injections and let @xmath221 be the monad such that for every @xmath207 , @xmath222 is the underlying set of the free algebra in @xmath95 on @xmath13 generators ( see ( * ? ? ?
* definition ii.10.9 ) and ( * ? ? ?
* vi.8 ) ) .
we have that @xmath49 is dual to @xmath48 , so we can consider @xmath185 . by using the duality between @xmath49 and @xmath48 , each coequational @xmath4theory can be indexed by @xmath127 and can be presented , up to isomorphism , as a family @xmath223 . from this , we present coequational @xmath4theories as operators @xmath206 on @xmath48 given by @xmath224 . then , we get a one to one correspondence between varieties of algebras in @xmath95 and operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath208 and it is a subalgebra of the complete atomic boolean algebra @xmath225 of subsets of @xmath222 .
b. @xmath210 is closed under derivatives with respect to the type @xmath91 .
that is , for every @xmath226 of arity @xmath227 , every @xmath228 , every @xmath229 , @xmath230 , and every @xmath211 we have that @xmath231 where @xmath232 is defined as @xmath233 @xmath234 .
that is , for every function symbol @xmath219 we get @xmath227 kinds of derivatives .
c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of @xmath221algebras @xmath235 and @xmath211 , we have that @xmath218 .
in fact , we have the following : a. condition i ) above follows from the fact that @xmath236 and @xmath237 .
b. condition ii ) above follows from lifting the duality between @xmath48 and @xmath49 to a duality between @xmath238 and @xmath26 .
in fact , every surjective @xmath221algebra morphism @xmath239 defines the injective morphism @xmath240 in @xmath26 which is defined as @xmath241 , @xmath242 , and from this we have : @xmath243 closure of @xmath210 under under derivatives with respect to the type @xmath91 follows from the fact that @xmath112 is a @xmath221algebra morphism .
in fact , for every @xmath242 we have : @xmath244 where the function @xmath245 is defined for every @xmath246 as @xmath247 . therefore , @xmath248 , i.e. , @xmath210 is closed under derivatives with respect to the type @xmath91 .
+ conversely , any @xmath249 closed under derivatives with respect to the type @xmath91 such that @xmath11 is a subalgebra of @xmath250 will define , by duality , the canonical surjective function @xmath251 where @xmath252 is defined as : @xmath253 where @xmath254 is the set of atoms of @xmath11 .
clearly , @xmath255 is an equivalence relation on @xmath222 since @xmath254 is a partition of @xmath222 .
we only need to show that @xmath255 is an @xmath91congruence on @xmath256.congruence on an algebra @xmath172 of type @xmath91 is an equivalence relation @xmath257 on @xmath81 such that for every @xmath219 of arity @xmath227 , every @xmath228 and @xmath258 , @xmath230 , the property @xmath259 implies @xmath260 .
* definition ii.5.1 ) ) ] in fact , let @xmath219 of arity @xmath227 , let @xmath228 , let @xmath229 , @xmath230 , and assume @xmath259 , i.e. , there exists @xmath261 such that @xmath262 , we have to show that @xmath263 . if @xmath264 then there exists @xmath265 such that @xmath266 which means that @xmath267 with @xmath268 by closure under derivatives with respect to the type @xmath91 .
therefore @xmath269 is an element in @xmath11 such that @xmath270 which contradicts the fact that @xmath81 is an atom .
this proves that @xmath263 , which means that @xmath271 is a surjective @xmath1algebra morphism . c. condition
iii ) above is the commutativity of the diagram in definition [ defcoeqth ] .
conversely , each operator @xmath206 on @xmath48 with the properties i ) , ii ) and iii ) above defines the coequational @xmath4theory @xmath272 where @xmath273 is the inclusion @xmath4coalgebra morphism .
note that conditions i ) and ii ) above are exactly the properties that @xmath210 is a @xmath4subcoalgebra of @xmath225 .
[ exvv1 ] from the previous general example we can provide details for the properties i ) , ii ) and iii ) in example [ exjan11 ] . in fact , for the case of monoids we have the type @xmath274 where @xmath68 is a nullary function symbol and @xmath66 is a binary function symbol .
we write @xmath275 for @xmath276 . by considering the variety @xmath95 of monoids
, we get the monad @xmath221 such that @xmath277 , where @xmath184 is the free monoid on @xmath13 .
then , we have : 1 .
properties i ) and iii ) in example [ exua ] trivially become properties i ) and iii ) in example [ exjan11 ] .
property ii ) in example [ exua ] does not give us any kind of derivatives for the nullary function symbol @xmath278 , but will give us the derivatives @xmath279 and @xmath280 for the binary function symbol @xmath281 , @xmath282 , which are defined for every @xmath215 as @xmath283 which are respectively the left and right derivatives of @xmath284 with respect to @xmath285 . in a similar way , from example [ exua ] , we get the following eilenberg type correspondences 1 . a one to one correspondence between varieties of semigroups and operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath208 and it is a subalgebra of the complete atomic boolean algebra @xmath286 of subsets of @xmath287 , i.e. , every element in @xmath210 is a language on @xmath13 not containing the empty word .
b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath132 then @xmath212 , where @xmath213 and @xmath214 , @xmath288 . c. @xmath206 is closed under morphic preimages
that is , for every @xmath216 , homomorphism of semigroups @xmath289 and @xmath211 , we have that @xmath218 .
2 . a one to one correspondence between varieties of groups and operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath208 and it is a subalgebra of the complete boolean algebra @xmath290 of subsets of the free group @xmath291 on @xmath13
. b. @xmath210 is closed under left and right derivatives and inverses .
that is , if @xmath211 and @xmath132 then @xmath292 , where @xmath213 , @xmath214 and @xmath293 , @xmath294 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of groups @xmath295 and @xmath211 , we have that @xmath218 .
3 . for a fixed monoid @xmath296 a one to one correspondence between varieties of @xmath297actions , i.e. , dynamical systems on @xmath297 , and operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath208 and it is a subalgebra of the complete atomic boolean algebra @xmath298 of subsets of @xmath299 .
b. @xmath210 is closed under translations .
that is , if @xmath211 and @xmath300 then @xmath301 , where @xmath302 , @xmath303 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of @xmath297actions @xmath304 ( i.e. , @xmath305 ) and @xmath211 , we have that @xmath218 .
4 . consider the type of algebras @xmath306 where @xmath66 is a binary operation and @xmath307 is a unary operation .
now , let @xmath1 be the free monad on @xmath48 for the algebras of type @xmath91 that satisfy the following equations : + m4cmm4cmm4 cm @xmath308 & @xmath309 & @xmath310 + @xmath311 , @xmath312 & @xmath313 & + + here @xmath275 is the product of @xmath314 and @xmath315 , in that order , and @xmath316 represents the infinite product @xmath317 .
hence , for every @xmath207 the algebra @xmath87 has as carrier set the set @xmath318 , where @xmath319 represents the set of all ultimately periodic sequences in @xmath320 , i.e. , every element in @xmath319 is of the form @xmath321 for some @xmath322 , and @xmath318 has the natural operations @xmath66 of concatenation and @xmath307 of `` infinite power '' .
+ in this case , we get a one to one correspondence between varieties of semigroups with infinite exponentiation and operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath208 and it is a subalgebra of the complete atomic boolean algebra @xmath323 of subsets of @xmath324
. b. @xmath210 is closed under left and right derivatives and infinite exponentiation .
that is , if @xmath211 and @xmath325 then @xmath326 , where @xmath327 , @xmath328 and @xmath329 , @xmath330 . c. @xmath206 is closed under morphic preimages
that is , for every @xmath216 , homomorphism of @xmath1algebras @xmath331 and @xmath211 , we have that @xmath218 . we can do a similar work as in example [ exua ] to get eilenberg type correspondences for varieties of ordered algebras for any given type .
[ exuoa ] let @xmath91 be a type of algebras where each function symbol @xmath219 has arity @xmath220 and let @xmath95 be a variety of ordered algebras of type @xmath91 . consider the case @xmath93 , @xmath94 discrete posets , @xmath89 surjections , @xmath90 embeddings and let @xmath221 be the monad such that for every @xmath332 , @xmath333 is the underlying poset of the free ordered algebra in @xmath95 on @xmath334 generators ( see , e.g. , ( * ? ? ?
* proposition 1 ) ) .
we have that @xmath51 is dual to @xmath50 , so we can consider @xmath335 , @xmath336 .
similar to example [ exua ] , by using the duality between @xmath50 and @xmath51 , each coequational @xmath4theory can be indexed by @xmath48 ( i.e. , we consider every object @xmath207 as the object @xmath337 , which is in @xmath74 ) and can be presented , up to isomorphism , as a family @xmath223 , then we present coequational @xmath4theories as operators @xmath206 on @xmath48 given by @xmath224 .
then , we get a one to one correspondence between varieties of ordered algebras in @xmath95 and operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath338 and it is a subalgebra of the algebraic completely distributive lattice @xmath339 of subsets of @xmath222 .
here @xmath340 is the two element chain .
b. @xmath210 is closed under derivatives with respect to the type @xmath91 .
that is , for every @xmath226 of arity @xmath227 , every @xmath228 , every @xmath229 , @xmath230 , and every @xmath211 we have that @xmath231 where @xmath232 is defined as @xmath233 @xmath234 .
that is , for every function symbol @xmath219 we get @xmath227 kinds of derivatives . c.
@xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of @xmath221algebras @xmath235 and @xmath211 , we have that @xmath218 . [ exvv3 ] from the previous example we can obtain eilenberg type correspondences for varieties of ordered semigroups , varieties of ordered monoids , varieties of ordered groups , and so on . for instance , for the case of varieties of ordered semigroups we can consider the type @xmath341 where @xmath66 is a binary function symbol and @xmath95 is the variety of ordered semigroups .
then we get a one to one correspondence between varieties of ordered semigroups and operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath338 and it is a subalgebra of the algebraic completely distributive lattice @xmath286 of subsets of @xmath287 , i.e. every element in @xmath210 is a language on @xmath13 not containing the empty word .
in particular , @xmath210 is closed under unions and intersections .
b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath132 then @xmath212 , where @xmath213 and @xmath214 , @xmath288 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of semigroups @xmath289 and @xmath211 , we have that @xmath218 . [ exvv4 ]
let @xmath7 be a finite field .
consider the case @xmath342 , @xmath89 surjections , and @xmath90 injections .
we have that @xmath53 is dual to @xmath52 , so we can consider @xmath343 .
for every set @xmath13 denote by @xmath344 the @xmath7vector space with basis @xmath13 .
consider the monad @xmath345 , where @xmath184 is the free monoid on @xmath13 .
then we get a one to one correspondence between varieties of @xmath7algebras and operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath346 and it is a subspace of the space @xmath347 where the topology on @xmath347 is the subspace topology of the product @xmath348
. b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath349 then @xmath350 , where @xmath351 and @xmath352 , @xmath353 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , @xmath7linear map @xmath354 and @xmath211 , we have that @xmath218 . [ exvv5 ] consider the case @xmath355 , @xmath94 free join semilattices , i.e. , @xmath356 , where @xmath357 is the set of all finite subsets of @xmath13 , @xmath89 surjections and @xmath90 injections .
we have that @xmath55 is dual to @xmath54 , so we can consider @xmath358 and @xmath359 .
let @xmath1 be the monad on @xmath54 such that @xmath360 is the free idempotent semiring on @xmath361 .
then we get a one to one correspondence between varieties of idempotent semirings and operators @xmath206 on @xmath48 such that for every @xmath207 : a. @xmath362 and it is a subspace of @xmath209 where the topology given on @xmath209 is the subspace topology of the product @xmath363 . in particular ,
@xmath210 is closed under union .
b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath132 then @xmath212 , where @xmath213 and @xmath214 , @xmath215 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , semiring homomorphism @xmath364 and @xmath211 , we have that @xmath365 , where @xmath366 and @xmath367 are defined as @xmath368 and @xmath369 .
note that the composite @xmath370 is the same as @xmath371 defined in @xcite .
the reason of the exponent @xmath372 and the use of @xmath373 is that we are using the isomorphism : @xmath374 see the appendix for more details .
note that eilenberg
type correspondences for varieties of @xmath7algebras and idempotent semirings can also be obtained from example [ exua ] .
this section is similar to the previous one with the restriction that all the algebras considered are finite .
we state a categorical version of birkhoff s theorem for finite @xmath1algebras and an abstract eilenberg type correspondence for pseudovarieties of @xmath1algebras .
we use the prefix ` pseudo ' to indicate that all the algebras considered are finite .
that is , a pseudovariety of @xmath1algebras is a variety of finite @xmath1algebras , which is a class of finite @xmath1algebras closed under homomorphic images , subalgebras and finite producs .
the birkhoff variety theorem for finite algebras has been previously proved to prove that a class of finite algebras of the same type is a pseudovariety , i.e. , it is closed under subalgebras , homomorphic images and finite products , if and only if it is defined by ` extended equations ' @xcite .
an ` extended equation ' is a concept that generalizes the concept of an equation and can be defined by using topological techniques or , alternatively , by implicit operations @xcite .
reiterman s proof for the birkhoff theorem for finite algebras involves topological methods in which the set of @xmath375ary implicit operations is the completion of the set of @xmath375ary terms @xcite .
a topological approach was also explored by banaschewski by using uniformities @xcite .
recently , in @xcite , profinite techniques were used to define the concept of profinite equations which are the kind of equations that define pseudovarieties of @xmath1algebras .
we provide a categorical version of the birkhoff theorem for finite algebras , theorem [ reitthm ] , which , under mild assumptions , establishes a one to
one correspondence between pseudovarieties of @xmath1algebras and pseudoequational @xmath1theories
. different versions of this theorem such as @xcite use topological approaches and/or profinite techniques . in the present paper , topological approaches and profinite techniques
are not used , thus avoiding constructions of certain limits and profinite completions , which gives us a better and basic understanding on how pseudovarieties are characterized . the main strategy
we follow to state and prove our theorem is that pseudovarieties of algebras are exactly directed unions of equational classes of finite algebras , which is a fact that was proved in @xcite .
the definition of pseudoequational @xmath1theories is based on the previous observation and the categorical dual of `` varieties of languages '' that was used by the author to derive an eilenberg type correspondence for @xmath1algebras @xcite . as in the previous section ,
the main purpose of this approach is to derive eilenberg
type correspondences for pseudovarieties of @xmath1algebras .
this is summarized in the following picture : ( ) at ( 0,-0.05 ) pseudovarieties ; ( ) at ( 0,-0.45 ) of @xmath1algebras ; ( 1.5,-0.25 ) ( 5.5,-0.25 ) ; ( ) at ( 3.5,-0.5 ) birkhoff s thm . for ; ( ) at ( 3.5,-0.9 ) finite @xmath1alg .
[ reitthm ] ; ( ) at ( 7,-0.05 ) pseudoequational ; ( ) at ( 7,-0.45 ) @xmath1theories ; ( 7,-0.75 )
( 7,-1.7 ) ; ( ) at ( 7.7,-1.2 ) duality ; ( ) at ( 7,-1.75 ) pseudocoequational ; ( ) at ( 7,-2.15 ) @xmath4theories ; ( 0,-0.75 ) ( 0,-1.95)(5.5,-1.95 ) ; ( ) at ( 2.5,-2.2 ) eilenberg type correspondence ; ( ) at ( 2.5,-2.6 ) proposition [ eilpvar ] ; throughout this section , we fix a complete concrete category @xmath2 such that its forgetful functor preserves epis , monos and products , a monad @xmath14 on @xmath2 , a full subcategory @xmath74 of @xmath2 and a factorization system @xmath61 on @xmath2 .
we make the following assumptions : 1 .
the factorization system @xmath61 is proper .
2 . for every @xmath75 , the free @xmath1algebra @xmath76 is _ projective with respect to @xmath59 in @xmath15_. that is , for every @xmath77 with @xmath75 and @xmath78 there exists @xmath79 such that @xmath376 .
3 . for every finite @xmath84
there exists @xmath85 and @xmath86 .
@xmath73 preserves morphisms in @xmath59 . in order to talk about finite algebras
, we assume that the category @xmath2 is a concrete category . that is , if @xmath377 is the forgetful functor for the concrete category @xmath2 , then an object @xmath378 is _ finite _ if @xmath379 is a finite set . similarly , an algebra @xmath84 is _ finite _ if its carrier object @xmath380 is finite .
the algebras of interest will be the objects @xmath57 of finite algebras in @xmath15 .
the factorization system @xmath61 on @xmath2 , which is lifted to @xmath15 by using ( b@xmath3811 ) and ( b@xmath3814 ) , allows us to define the concept of homomorphic image and subalgebra . in this case
, the requirement of the forgetful functor @xmath382 preserving epis , monos and products , will give us the property that subalgebras , homomorphic images and finite products of finite algebras are also finite .
the purpose of the subcategory @xmath74 is that the objects from which `` variables '' for the equations are considered are objects in @xmath74 .
assumption ( b@xmath3813 ) guarantees that every algebra is the homomorphic image of a free one with object of generators in @xmath74 . to obtain birkhoff s theorem for finite algebras we can consider @xmath383 , @xmath94 finite sets , @xmath89 surjections , @xmath90 injections , and @xmath1 to be the term monad for a given type of algebras @xmath91 , i.e. , @xmath92 , the set of terms of type @xmath91 on the set of variables @xmath13 ( see example [ excleqth ] ) .
another important example will be given by @xmath93 and @xmath74 to be the full subcategory of finite discrete posets ( as before , we do not want the `` variables '' to be ordered ) .
now , we will define the main concepts needed to state our categorical birkhoff s theorem for finite @xmath1algebras .
[ defpet ] let @xmath2 be a complete concrete category such that its forgetful functor preserves epis , @xmath1 a monad on @xmath2 , @xmath74 a full subcategory of @xmath74 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) and ( b@xmath3814 ) .
a _ pseudoequational @xmath1theory on @xmath74 _ is an operator @xmath384 on @xmath74 such that for every @xmath75 , @xmath385 is a nonempty collection of @xmath1algebra morphisms in @xmath59 with domain @xmath87 and finite codomain and : a. for every finite set @xmath100 and @xmath386 , @xmath102 , there exists @xmath387 such that every @xmath388 factors through @xmath62 , @xmath102 .
b. for every @xmath389 with codomain @xmath172 and every @xmath1algebra morphism @xmath390 with domain @xmath172 we have that @xmath391 . c. for every @xmath392 , @xmath387 and @xmath393 we have that @xmath394 where @xmath395 is the factorization of @xmath396 .
pseudovarieties of algebras are exactly directed unions of equational classes of finite algebras @xcite . with this in mind , we can give an intuition of the previous definition .
in fact , for each object @xmath75 of variables every morphism @xmath387 represents a set of equations on @xmath13 , namely @xmath397 , which can be equivalently given by a @xmath1algebra morphism in @xmath59 with domain @xmath87 .
condition i ) says that the set of all the equations on a fixed @xmath13 is a directed set , i.e. , for every set of equations @xmath386 , @xmath102 , with @xmath100 finite , there is an upper bound @xmath387 .
here @xmath62 is an upper bound of @xmath398 if every @xmath388 factors through @xmath62 . condition iii ) says that all the equations considered are preserved under any substitution @xmath393 of variables in @xmath28 by terms in @xmath82 , this condition is related to the commutativity of the diagram given in definition [ defeqth ] .
condition ii ) is needed for uniqueness of the pseudoequational theory defining a given pseudovariety of algebras .
in fact , two directed unions of equational classes of finite algebras can give us the same pseudovariety , but if we put the requirement of being downward closed , which is the requirement in condition ii ) , then we get uniqueness .
given an algebra @xmath399 , we say that @xmath172 _ satisfies _ @xmath384 , denoted as @xmath400 , if for every @xmath75 and @xmath401 we have that @xmath62 factors through some morphism in @xmath385 .
we denote by @xmath402 the _ finite models _ of @xmath384 , that is : @xmath403 a class @xmath95 of finite @xmath1algebras is _ defined _ by @xmath384 if @xmath404 .
let @xmath95 be a class of algebras in @xmath57 .
we say that @xmath95 _ is closed under @xmath59quotients _ if @xmath405 for every @xmath78 with @xmath97 .
we say that @xmath95 _ is closed under @xmath60subalgebras _ if @xmath405 for every @xmath98 with @xmath97 .
we say that @xmath95 _ is closed under finite products _ if @xmath99 for every finite set @xmath100 such that @xmath101 , @xmath102 .
let @xmath2 be a complete concrete category , @xmath1 a monad on @xmath2 and @xmath61 a factorization system on @xmath2 .
a class @xmath95 of finite algebras in @xmath15 is called a _
pseudovariety of @xmath1algebras _ if it is closed under @xmath59quotients , @xmath60subalgebras and finite products .
consider the setting @xmath383 , @xmath94 finite sets , @xmath89 surjections , @xmath90 injections , and @xmath1 to be the term monad for a given type of algebras @xmath91 .
then we have that equational classes of finite algebras are examples of pseudovarieties of @xmath1algebras . for example , finite semigroups , finite monoids , finite groups , finite vector spaces , finite boolean algebras , finite lattices , and so on . in @xcite ,
some non equational examples of pseudovarieties are shown such as : 1 .
the finite commutative monoids satisfying some identity @xmath406 , @xmath407 , 2 .
the finite cancellation monoids , 3 .
the finite abelian @xmath408groups , for a given prime number @xmath408 , and 4 .
the finite products of finite fields of a given prime characteristic .
each of those pseudovarieties is not equational .
in fact , every equation satisfied in the given pseudovariety is also satisfied in the larger pseudovariety , i.e. , the pseudovariety of all commutative monoids for ( 1 ) , the pseudovariety of all monoids for ( 2 ) , the pseudovariety of all abelian groups for ( 3 ) , and the pseudovariety of all commutative rings with unit of a given prime characteristic for ( 4 ) .
now we can formulate our categorical birkhoff s theorem for finite @xmath1algebras as follows .
[ reitthm ] let @xmath2 be a complete concrete category such that its forgetful functor preserves epis , monos and products , @xmath1 a monad on @xmath2 , @xmath74 a full subcategory of @xmath74 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) to ( b@xmath3814 ) .
then a class @xmath95 of finite @xmath1algebras is a pseudovariety of @xmath1algebras if and only if is defined by a pseudoequational @xmath1theory on @xmath74 .
additionally , pseudovarieties of @xmath1algebras are in one to one correspondence with pseudoequational @xmath1theories on @xmath74 .
now we derive birkhoff s theorem for pseudovarieties of ( ordered ) algebras for a given type , then show an example of a particular pseudovariety of algebras with its defining pseudoequational @xmath1theory and finish this subsection by deriving eilenberg s theorem ( * ? ? ?
* theorem 34 ) to show a one to one correspondence between pseudovarieties of monoids and pseudovarieties of languages .
consider the case @xmath383 , @xmath94 finite sets , @xmath89 surjections , @xmath90 injections and , for a given type of algebras @xmath91 , let @xmath126 be the term monad for @xmath91 .
then , by theorem [ reitthm ] , a class of algebras of type @xmath91 is a pseudovariety if and only if it is defined by a pseudoequational @xmath126theory .
consider the case @xmath93 , @xmath94 finite discrete posets , @xmath89 surjections , @xmath90 embeddings and , for a given type of algebras @xmath91 , let @xmath126 be the monad on @xmath50 defined in example [ exineqth ] .
then , by theorem [ reitthm ] , a class of ordered algebras of type @xmath91 is a pseudovariety if and only if it is defined by a pseudoequational @xmath126theory .
consider the case @xmath383 , @xmath94 finite sets , @xmath1 the monad given by @xmath183 , where @xmath184 is the free monoid on @xmath13 , @xmath89 surjections , and @xmath90 injections .
we have that conditions ( b@xmath3811 ) to ( b@xmath3814 ) are fullfilled . in this case
, we have that @xmath15 is the category of monoids . to describe the pseudovariety of all commutative monoids satisfying some identity @xmath406 , @xmath407 , we define @xmath384 on @xmath74 as follows : 1 . for every @xmath75 , and @xmath409 , we define the surjective homomorphism of monoids @xmath410 , where @xmath411 is the free commutative monoid on @xmath13 that satisfies the identity @xmath406 .
that is , @xmath412 where @xmath413 is the zero function , i.e. , @xmath414 for every @xmath132 , and @xmath66 is defined on @xmath415 as @xmath416 .
@xmath417 is defined on the set of generators @xmath13 as @xmath418 , where @xmath419 and @xmath420 for @xmath421 .
define @xmath385 as : @xmath422 we have then that @xmath384 is a pseudoequational @xmath1theory and @xmath423 is the pseudovariety of all finite commutative monoids that satisfy some identity @xmath406 , @xmath407 . in the next example
we derive eilenberg s variety theorem ( * ? ? ?
* theorem 3.4 . ) . given a finite set @xmath5 , i.e. , an _ alphabet _ , a _ language _
@xmath284 on @xmath5 is a subset @xmath284 of @xmath6 , i.e. , a collection of words with letters in @xmath5 .
we identify a language @xmath284 on @xmath5 by its characteristic function @xmath424 .
a language @xmath284 on @xmath5 is _ recognizable _ if there exists a finite monoid @xmath172 , a homomorphism of monoids @xmath425 and a function @xmath426 such that @xmath427 .
we denote by @xmath428 the boolean algebra of all recognizable languages on @xmath5 .
pseudovariety of languages _ is an operator @xmath206 such that for every finite set @xmath5 we have : a. @xmath429 is a subalgebra of the boolean algebra @xmath428 , b. @xmath429 is closed under left and right derivatives .
that is , @xmath430 for every @xmath431 and @xmath432 , and c. @xmath206 is closed under morphic preimages .
that is , for every alphabet @xmath433 , homomorphism of monoids @xmath434 and @xmath431
, we have that @xmath435 . eilenberg s variety theorem ( * ? ? ? * theorem 34 ) says that there is a one to
one correspondence between pseudovarieties of monoids and pseudovarieties of languages .
this theorem is derived from theorem [ reitthm ] as follows .
[ exeil ] consider the setting as in the previous example , i.e. , @xmath383 , @xmath94 finite sets , @xmath1 the monad given by @xmath183 , where @xmath184 is the free monoid on @xmath13 , @xmath89 surjections , and @xmath90 injections .
then , we have a one to one correspondence between pseudovarieties of monoids , i.e. , pseudovarieties of @xmath1algebras , and pseudoequational @xmath1theories on @xmath74 .
now , we have that pseudoequational @xmath1theories on @xmath74 are in one to
one correspondence with pseudovarieties of languages .
in fact , every pseudoequational @xmath1theory @xmath384 on @xmath74 defines the pseudovariety of languages @xmath436 defined as @xmath437 , and every pseudovariety of languages @xmath206 defines the pseudoequational @xmath1theory @xmath438 on @xmath74 such that @xmath439 is the collection of all @xmath1algebra morphisms @xmath64 with domain @xmath87 and finite codomain such that @xmath440 , @xmath75 .
furthermore , this correspondence is bijective , that is , for every pseudoequational @xmath1theory @xmath384 on @xmath74 and every pseudovariety of languages @xmath206 we have that @xmath441 and @xmath442 ( see example [ exuaf ] for more details ) . as we saw in example [ exeil ] , we can derive eilenberg
type correspondences for pseudovarieties of @xmath1algebras from birkhoff s theorem for finite @xmath1algebras , theorem [ reitthm ] .
type correspondences for pseudovariaties of @xmath1algebras are exactly one to one correspondences between pseudovarieties of @xmath1algebras and duals of pseudoequational @xmath1theories . by dualizing the definition of a pseudoequational @xmath1theory we get the following .
[ defpcoet ] let @xmath3 be a concrete category such that its forgetful functor preserves monos , @xmath443 a comonad on @xmath3 , @xmath61 a factorization system on @xmath3 and @xmath193 a full subcategory of @xmath3 .
assume ( b@xmath3811 ) and that @xmath80 preserves the morphisms in @xmath60 .
pseudocoequational @xmath4theory on @xmath193 _ is an operator @xmath444 on @xmath193 such that for every @xmath445 , @xmath446 is a nonempty collection of @xmath4coalgebra morphisms in @xmath60 with codomain @xmath447 and finite domain and : a. for every finite set @xmath100 and @xmath448 , @xmath102 , there exists @xmath449 such that every @xmath388 factors through @xmath62 , @xmath102 .
b. for every @xmath450 with domain @xmath172 and every @xmath4coalgebra morphism @xmath451 with codomain @xmath172 we have that @xmath452 . c. for every @xmath453 , @xmath449 and @xmath454 we have that @xmath455 where @xmath456 is the factorization of @xmath457 . with the previous definition , theorem [ reitthm ] and duality
, we have the following : [ eilpvar ] let @xmath2 be a complete concrete category such that its forgetful functor preserves epis , monos and products , @xmath1 a monad on @xmath2 , @xmath74 a full subcategory of @xmath74 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) to ( b@xmath3814 ) .
let @xmath3 be a category that is dual to @xmath2 , let @xmath193 be dual of @xmath74 and @xmath4 be the comonad on @xmath3 that is dual to the monad @xmath1 on @xmath2 which is defined as in proposition [ monadtocomonad ] .
then there is a one to
one correspondence between pseudovarieties of @xmath1algebras and pseudocoequational @xmath4theories on @xmath193 .
we will consider the same settings given in the examples in subsection [ seceilv ] to obtain the following eilenberg type correspondences for pseudovarieties of @xmath1algebras . in all of them
, we only need to change the category @xmath74 and condition i ) for the operators @xmath206 ( see appendix for their details ) .
[ exuaf ] in this example we obtain an eilenberg type correspondence for any pseudovariety of algebras of any given type @xmath91 , where each of the function symbols in @xmath91 has finite arity .
let @xmath91 be a type of algebras where each function symbol @xmath219 has arity @xmath220 and let @xmath95 be a variety of algebras for of type @xmath91 . consider the case @xmath383 , @xmath458 , @xmath89 surjections , @xmath90 injections and let @xmath221 be the monad such that for every @xmath207 , @xmath222 is the underlying set of the free algebra in @xmath95 on @xmath13 generators ( see ( * ? ? ?
* definition ii.10.9 ) and ( * ? ? ?
* vi.8 ) ) .
we have that @xmath49 is dual to @xmath48 , so we can consider @xmath459 and @xmath460 . in this case , we get a one to one correspondence between pseudovarieties of algebras in @xmath95 and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a boolean algebra and it is a subalgebra of the complete atomic boolean algebra @xmath225 of subsets of @xmath222 such that for every @xmath211 there exists a finite algebra @xmath172 in @xmath95 , a morphism @xmath463 and @xmath464 such that @xmath465 .
b. @xmath210 is closed under derivatives with respect to the type @xmath91 .
that is , for every @xmath226 of arity @xmath227 , every @xmath228 , every @xmath229 , @xmath230 , and every @xmath211 we have that @xmath231 where @xmath232 is defined as @xmath233 @xmath234 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath466 , homomorphism of @xmath221algebras @xmath235 and @xmath211 , we have that @xmath218 .
in fact , let @xmath384 be a pseudoequational @xmath221theory on @xmath461 and let @xmath206 be an operator on @xmath461 satisfying the three properties above .
then : a. define the operator @xmath436 on @xmath461 as @xmath437 .
we claim that @xmath436 satisfies the three properties above .
in fact , as the family @xmath385 is directed in the sense of definition [ defpet ] i ) , then the union @xmath467 is a directed union of finite objects in @xmath49 which is a boolean subalgebra of @xmath225 .
as each @xmath389 has as codomain a finite algebra in @xmath95 then @xmath468 is a subset of @xmath225 which is closed under derivatives with respect to the type @xmath91 ( see example [ exua ] ) .
the previous argument shows that @xmath436 satisfies properties i ) and ii ) above .
now , closure under morphic preimages follows from property iii ) in definition [ defpet ] .
therefore , @xmath436 satisfies the three properties above .
b. define the operator @xmath438 on @xmath461 such that @xmath439 is the collection of all @xmath221algebra morphisms @xmath64 with domain @xmath256 and finite codomain such that @xmath440 .
we claim that @xmath438 is a pseudoequational @xmath221theory .
in fact , we have that @xmath439 is nonempty since @xmath469 , where @xmath470 is the one element @xmath221algebra . by definition
, we have that @xmath439 satisfies property ii ) in definition [ defpet ] , and , it also satisfies property iii ) in definition [ defpet ] since @xmath206 is closed under morphic preimages .
now , consider a family @xmath471 in @xmath439 with @xmath100 finite such that @xmath472 , we need to find a morphism @xmath473 such that every @xmath474 factors through @xmath68 .
in fact , let @xmath172 be the product of @xmath475 with projections @xmath476 , then , by the universal property of @xmath172 there exists a @xmath221algebra morphism @xmath477 such that @xmath478 , for every @xmath102 .
let @xmath479 be the factorization of @xmath62 in @xmath480 .
we claim that @xmath481 is a morphism in @xmath439 such that every @xmath474 factors through @xmath68 .
clearly , from the construction above , each @xmath474 factors through @xmath481 .
now , let s prove that @xmath440 .
in fact , let @xmath482 be the codomain of @xmath481 and let @xmath483 .
we have to prove that @xmath484 which follows from the following straightforward identity : @xmath485 where @xmath486 is the set @xmath487 ( i.e. , we express the subset @xmath69 of @xmath11 as the union of its points and each point @xmath488 is represented as @xmath489 ) .
now , for every @xmath488 and @xmath102 the composition @xmath490 belongs to @xmath210 since @xmath491 .
as @xmath11 and @xmath100 are finite then @xmath484 because @xmath210 is a boolean algebra .
we have that @xmath441 .
in fact , for every @xmath462 the inclusion @xmath492 is obvious .
now , to prove that @xmath493 , let @xmath494 with finite codomain such that @xmath495 , i.e. , @xmath496 .
then the previous inclusion means that for every @xmath497 there exists @xmath498 and @xmath176 such that @xmath499 .
as @xmath500 is finite , then there exists @xmath389 such that each @xmath501 factors through @xmath68 .
we will prove that @xmath502 factors through @xmath389 which will imply that @xmath503 , since @xmath384 is a pseudoequational @xmath221theory .
it is enough to show that @xmath504 .
in fact , assume that @xmath505 and define @xmath506 as @xmath507 iff @xmath508 . then , as @xmath509 factors through @xmath68 we have that @xmath510 which implies @xmath511 .
but @xmath512 , which implies that @xmath513 , i.e. , @xmath514 , but the later equality means that @xmath515 by definition of @xmath516 , i.e. , @xmath517 as desired .
d. we have that @xmath442 .
in fact , for every @xmath462 the inclusion @xmath518 is obvious .
now , to prove @xmath519 we need to find for every @xmath211 a surjective homomorphism @xmath520 with @xmath97 such that @xmath521 .
in fact , for @xmath211 let @xmath522 be a homomorphism with @xmath405 and @xmath523 such that @xmath524 , this can be done by property i ) above .
let @xmath525 be the subset of @xmath225 obtained from @xmath526 which is closed under boolean combinations and derivatives with respect to the type @xmath91 .
we show that @xmath527 , that is , we show that @xmath525 is a finite object in @xmath49 that is closed under derivatives with respect to the type @xmath91 .
in fact , @xmath528 is such that @xmath529 , which implies that @xmath525 is a finite boolean algebra , i.e. , an object in @xmath58 . by construction of @xmath525
we have that @xmath530 since @xmath206 satisfies properties i ) and ii ) above .
now , let @xmath531 be the inclusion morphism , then by duality we have that the dual morphism @xmath68 in @xmath480 of @xmath532 is such that @xmath521 ( in fact , @xmath533 ) . note that the codomain of @xmath68 is in @xmath95 since it is an @xmath59quotient of @xmath405 .
note that , for every `` language '' @xmath534 , the object @xmath525 in d ) above is the @xmath4subcoalgebra of @xmath225 generated by @xmath284 which implies , by duality , that its dual is the syntactic algebra @xmath535 of @xmath284 .
additionally , by using duality and the construction of @xmath525 , we have that every `` language '' in @xmath468 ( i.e. , recognized by the syntactic algebra of @xmath284 ) is a boolean combination of derivatives of @xmath284 , where @xmath68 is the dual of the inclusion @xmath531 .
[ expvv1 ] from the previous example , we get the following eilenberg type correspondences : 1 .
* theorem 34 ) a one to one correspondence between pseudovarieties of monoids and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a boolean subalgebra of @xmath209 such that for every @xmath211 there exists a finite monoid @xmath297 , a homomorphism @xmath536 and @xmath537 such that @xmath427 , i.e. , @xmath284 is a recognizable language on @xmath13 .
b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath132 then @xmath212 , where @xmath213 and @xmath214 , @xmath215 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of monoids @xmath217 and @xmath211 , we have that @xmath218 .
* theorem 34s ) a one to one correspondence between pseudovarieties of semigroups and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a boolean subalgebra of @xmath286 such that for every @xmath211 there exists a finite semigroup @xmath482 , a homomorphism @xmath538 and @xmath539 such that @xmath427 , i.e. , @xmath284 is a recognizable language on @xmath13 not containing the empty word .
b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath132 then @xmath212 , where @xmath213 and @xmath214 , @xmath288 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of semigroups @xmath289 and @xmath211 , we have that @xmath218 .
3 . a one to one correspondence between pseudovarieties of groups and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a boolean subalgebra of @xmath290 such that for every @xmath211 there exists a finite group @xmath540 , a homomorphism @xmath541 and @xmath542 such that @xmath427
. b. @xmath210 is closed under left and right derivatives and inverses .
that is , if @xmath211 and @xmath132 then @xmath292 , where @xmath213 , @xmath214 and @xmath293 , @xmath294 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of groups @xmath295 and @xmath211 , we have that @xmath218 .
4 . for a fixed monoid @xmath296 , a one to one correspondence between pseudovarieties of @xmath297actions ,
i.e. , dynamical systems on @xmath297 , and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a boolean subalgebra of @xmath298 such that for every @xmath211 there exists a finite @xmath297action @xmath482 , a homomorphism @xmath538 and @xmath539 such that @xmath427 .
b. @xmath210 is closed under translations .
that is , if @xmath211 and @xmath300 then @xmath301 , where @xmath302 , @xmath303 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of @xmath297actions @xmath304 ( i.e. , @xmath543 ) and @xmath211 , we have that @xmath218 .
@xcite ) a one to one correspondence between pseudovarieties of semigroups with infinite exponentiation and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a boolean subalgebra of @xmath323 such that for every @xmath211 there exists a finite semigroup with infinite exponentiation @xmath482 , a homomorphism @xmath538 and @xmath539 such that @xmath427
. b. @xmath210 is closed under left and right derivatives and infinite exponentiation .
that is , if @xmath211 and @xmath325 then @xmath326 , where @xmath327 , @xmath328 and @xmath329 , @xmath330 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of @xmath1algebras @xmath331 and @xmath211 , we have that @xmath218 . in the next example we obtain an eilenberg type correspondence for any variety of ordered algebras for a given type @xmath91 such that each function symbol in @xmath91 has a finite arity . [ exuoaf ]
let @xmath91 be a type of algebras where each function symbol @xmath219 has arity @xmath220 and let @xmath95 be a variety of ordered algebras of type @xmath91 . consider the case @xmath93 , @xmath94 finite discrete posets , @xmath89 surjections , @xmath90 embeddings and let @xmath221 be the monad such that for every @xmath332 , @xmath333 is the underlying poset of the free ordered algebra in @xmath95 on @xmath334 generators ( see ( * ? ? ?
* proposition 1 ) ) .
we have that @xmath51 is dual to @xmath50 , so we can consider @xmath335 , @xmath460 . in this case , we get a one to one correspondence between pseudovarieties of ordered algebras in @xmath95 and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a distributive sublattice of @xmath339 of subsets of @xmath222 such that for every @xmath211 there exists a finite ordered algebra @xmath172 in @xmath95 , a morphism @xmath463 and @xmath544 such that @xmath465 . here
@xmath340 is the two element chain .
b. @xmath210 is closed under derivatives with respect to the type @xmath91 .
that is , for every @xmath226 of arity @xmath227 , every @xmath228 , every @xmath229 , @xmath230 , and every @xmath211 we have that @xmath231 where @xmath232 is defined as @xmath233 @xmath234 .
that is , for every function symbol @xmath219 we get @xmath227 kinds of derivatives . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of @xmath221algebras @xmath235 and @xmath211 , we have that @xmath218 . [ expvv3 ] from the previous example we can obtain eilenberg type correspondences for pseudovarieties of ordered semigroups , pseudovarieties of ordered monoids , pseudovarieties of ordered groups , and so on . for instance , for the case of pseudovarieties of ordered monoids we can consider the type @xmath274 where @xmath68 is a nullary function symbol , @xmath66 is a binary function symbol and @xmath95 is the variety of ordered monoids . then we get a one to one correspondence between pseudovarieties of ordered monoids and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a distributive sublattice of the distributive lattice @xmath209 of subsets of @xmath184 , i.e. , every element in @xmath210 is a language on @xmath13 , such that every @xmath211 is a regular language
. b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath132 then @xmath212 , where @xmath213 and @xmath214 , @xmath215 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , homomorphism of monoids @xmath217 and @xmath211 , we have that @xmath218 . [ expvv4 ]
let @xmath7 be a finite field .
consider the case @xmath545 , @xmath94 finite @xmath7vector spaces , @xmath89 surjections and @xmath90 injections .
we have that @xmath53 is dual to @xmath52 , so we can consider @xmath546 and @xmath547 finite @xmath7vector spaces .
for every set @xmath13 denote by @xmath344 the @xmath7vector space with basis @xmath13 .
consider the monad @xmath345 , where @xmath184 is the free monoid on @xmath13 .
then we get a one to one correspondence between pseudovarieties of @xmath7algebras and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a @xmath7vector space which is a subspace of @xmath347 such that every element @xmath11 in @xmath210 is a _ recognizable series _ on @xmath13 , i.e. , there exists a @xmath7algebra morphism @xmath548 , with @xmath172 finite , and @xmath549 such that @xmath550
. b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath349 then @xmath350 , where @xmath351 and @xmath352 , @xmath353 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , @xmath7linear map @xmath354 and @xmath211 , we have that @xmath218 . [ expvv5 ] consider the case @xmath355 , @xmath94 finite free join semilattices , i.e. , @xmath551 , where @xmath552 is the powerset operator , @xmath89 surjections and @xmath90 injections .
we have that @xmath55 is dual to @xmath54 , so we can consider @xmath358 and @xmath553 .
let @xmath1 be the monad on @xmath54 such that @xmath360 is the free idempotent semiring on @xmath361 .
then we get a one to one correspondence between pseudovarieties of idempotent semirings and operators @xmath206 on @xmath461 such that for every @xmath462 : a. @xmath210 is a join subsemilattice of @xmath209 such that every @xmath211 is a regular language . in particular ,
@xmath210 is closed under union .
b. @xmath210 is closed under left and right derivatives .
that is , if @xmath211 and @xmath132 then @xmath212 , where @xmath213 and @xmath214 , @xmath215 . c. @xmath206 is closed under morphic preimages .
that is , for every @xmath216 , semiring homomorphism @xmath364 and @xmath211 , we have that @xmath365 ( see example [ exvv5 ] ) . note that eilenberg type correspondences for pseudovarieties of @xmath7algebras and idempotent semirings can also be obtained from example [ exuaf ] .
in this section , we provide abstract versions of local eilenberg type correspondences for local ( pseudo)varieties of @xmath1algebras .
local eilenberg type correspondences have been studied in @xcite .
the main idea of local eilenberg type correspondences is to work with a fixed alphabet , which in our notation reduces to consider the case in which the category @xmath74 has only one object , say @xmath13 . in order to do this , the kind of algebras considered in this local version are algebras that are generated by the object @xmath13 in the following sense .
let @xmath2 be a category , @xmath1 a monad on @xmath2 , @xmath61 a factorization system on @xmath2 and @xmath378 .
an algebra @xmath84 is @xmath13generated if @xmath554 is nonempty .
we have that @xmath59quotients of @xmath13generated @xmath1algebras are @xmath13generated , but this property does not hold in general for @xmath60subalgebras and products . thus , we will restrict our attention to @xmath13generated @xmath60subalgebras , i.e. , @xmath60subalgebras that are @xmath13generated , and subdirect products .
the latter are defined as follows .
let @xmath2 be a complete category , @xmath1 a monad on @xmath2 , @xmath61 a proper factorization system on @xmath2 such that @xmath73 preserves the morphisms in @xmath59 .
let @xmath378 and let @xmath555 be an @xmath13generated @xmath1algebra with @xmath556 , @xmath102 .
we define the _ subdirect product _ of the family @xmath557 as the @xmath13generated @xmath60subalgebra @xmath482 of @xmath558 described in the following commutative diagram : \(a ) at ( 0,1.5 ) @xmath82 ; ( b ) at ( 0,0 ) @xmath559 ; ( c ) at ( 2.5,0 ) @xmath560 ; ( d ) at ( -2.5,0 ) @xmath11 ; ( a ) edge [ ] node [ right ] @xmath561 ( c ) edge [ ] node [ left ] @xmath562 ( d ) ; ( b ) edge [ ] node [ below ] @xmath563 ( c ) ( a ) edge [ ] node [ right ] @xmath68 ( b ) ; ( d ) edge [ ] node [ below ] @xmath564 ( b ) ; where @xmath68 is obtained from the morphisms @xmath561 , @xmath565 , and the universal property of the product @xmath558 and @xmath566 is the factorization of @xmath68 .
we say that the subdirect product @xmath482 defined above is _ finite _ if @xmath100 is a finite set . to obtain local versions of eilenberg type correspondences , the concept of ( pseudo)variety used is : classes of ( finite ) @xmath13generated @xmath1algebras closed under @xmath59quotients , @xmath13generated @xmath60subalgebras and ( finite ) subdirect products .
we state the two corresponding local versions in the rest of this section .
proofs are made in a similar way by using local versions of birkhoff s theorem for ( finite ) @xmath1algebras . in this subsection , we provide eilenberg type correspondences for local varieties of @xmath1algebras . for this purpose , as in section [ secbirk ] , we first provide a local version of birkhoff s theorem .
we fix a complete category @xmath2 , a monad @xmath14 on @xmath2 , @xmath61 a factorization system on @xmath2 and @xmath378 .
we will use the following assumptions : 1 .
the factorization system @xmath61 is proper .
the free @xmath1algebra @xmath76 is _ projective with respect to @xmath59 in @xmath15_. 3 .
@xmath73 preserves morphisms in @xmath59 .
4 . there is , up to isomorphism , only a set of @xmath1algebra morphisms in @xmath59 with domain @xmath87 .
let @xmath2 be a complete category , @xmath1 a monad on @xmath2 , and @xmath61 a factorization system on @xmath2 .
assume ( b1 ) and ( b3 ) .
let @xmath378 .
a class @xmath95 of @xmath13generated @xmath1algebras is a _ local variety of @xmath13generated @xmath1algebras _ if it is closed under @xmath59quotients , @xmath13generated @xmath60subalgebras and subdirect products .
[ defleqth ] let @xmath2 be a category , @xmath1 a monad on @xmath2 , @xmath378 and @xmath61 a factorization system on @xmath2 .
a _ local equational @xmath1theory on @xmath13 _ is a @xmath1algebra morphism @xmath567 in @xmath59 such that for any @xmath568 there exists @xmath569 such that the following diagram commutes : \(a ) at ( 2,1.5 ) @xmath82 ; ( b ) at ( 2,0 ) @xmath110 ; ( c ) at ( 0,1.5 ) @xmath82 ; ( d ) at ( 0,0 ) @xmath110 ; ( a ) edge [ ] node [ right ] @xmath112 ( b ) ( c ) edge [ ] node [ left ] @xmath112 ( d ) ; ( c ) edge [ ] node [ above ] @xmath113 ( a ) ; ( d ) edge [ ] node [ below ] @xmath114 ( b ) ; note that , in the setting of example [ excleqth ] , a local equational @xmath1theory @xmath567 on @xmath13 is characterized , up to isomorphism , by its kernel @xmath570 . in this case
, the property of @xmath112 being a local equational @xmath1theory is exactly the property of @xmath115 being a fully invariant congruence of @xmath87 ( * ? ? ?
* definition ii.14.1 ) .
this generalizes the definition of an equational theory over @xmath13 in ( * ? ? ?
* definition ii.14.9 ) to a categorical level . given a local equational @xmath1theory @xmath567 on @xmath13 and an @xmath13generated @xmath1algebra @xmath172
, we say that @xmath172 _ satisfies _ @xmath112 , denoted as @xmath571 , if every morphism @xmath178 factors through @xmath112 .
we denote by @xmath572 the _ @xmath13generated models of _ @xmath112 , that is : @xmath573 a class @xmath95 of @xmath13generated @xmath1algebras is _ defined _ by @xmath112 if @xmath574 .
[ lbirkthm ] let @xmath2 be a complete category , @xmath1 a monad on @xmath2 , @xmath378 and @xmath61 a factorization system on @xmath2 .
assume ( b1 ) to ( b4 ) .
then a class @xmath95 of @xmath13generated @xmath1algebras is a local variety of @xmath13generated @xmath1algebras if and only if is defined by a local equational @xmath1theory on @xmath13 .
additionally , local varieties of @xmath13generated @xmath1algebras are in one to
one correspondence with local equational @xmath1theories on @xmath13 .
[ deflcoeqth ]
let @xmath3 be a category , @xmath4 a comonad on @xmath3 , @xmath575 and @xmath61 a factorization system on @xmath3 .
a _ local coequational @xmath4theory on @xmath28 _ is a @xmath4coalgebra morphism @xmath576 in @xmath60 such that for any @xmath577 there exists @xmath578 such that the following diagram commutes : \(a ) at ( 2,0 ) @xmath195 ; ( b ) at ( 2,1.5 ) @xmath196 ; ( c ) at ( 0,0 ) @xmath195 ; ( d ) at ( 0,1.5 ) @xmath196 ; ( a ) edge [ ] node [ right ] @xmath199 ( b ) ( c ) edge [ ] node [ left ] @xmath199 ( d ) ; ( d ) edge [ ] node [ above ] @xmath113 ( b ) ; ( c ) edge [ ] node [ below ] @xmath114 ( a ) ; with the previous definition , theorem [ lbirkthm ] and duality , we have the following . [ leilvar ]
let @xmath2 be a complete category , @xmath1 a monad on @xmath2 , @xmath61 a factorization system on @xmath2 and @xmath378 .
assume ( b1 ) to ( b4 ) .
let @xmath3 be a category that is dual to @xmath2 , @xmath28 the corresponding dual object of @xmath13 and let @xmath4 be the comonad on @xmath3 that is dual to @xmath1 which is defined as in proposition [ monadtocomonad ] .
then there is a one to
one correspondence between local varieties of @xmath13generated @xmath1algebras and local coequational @xmath4theories on @xmath28 .
[ exlvv ] by fixing an object @xmath75 , we can get corresponding local versions of the eilenberg type correspondences showed in subsection [ seceilv ] for varieties of @xmath13generated @xmath1algebras .
for example , the local version of example [ exjan11 ] reads as follows : there is a one to
one correspondence between varieties of @xmath13generated monoids and subalgebras @xmath579 of the complete atomic boolean algebra @xmath209 such that : a. @xmath11 is closed under left and right derivatives .
that is , for every @xmath580 and @xmath132 , @xmath581 .
b. @xmath11 is closed under morphic preimages .
that is , for every homomorphism of monoids @xmath582 and @xmath580 , we have that @xmath583 . in this subsection , we provide an abstract eilenberg type correspondence for local pseudovarieties of @xmath1algebras . in order to do this ,
we first provide a local version of birkhoff s theorem for finite @xmath1algebras .
we fix a complete concrete category @xmath2 such that its forgetful functor preserves epis , monos and products , a monad @xmath14 on @xmath2 , @xmath378 and a factorization system @xmath61 on @xmath2 .
we will need the following assumptions : 1 .
the factorization system @xmath61 is proper .
the free @xmath1algebra @xmath76 is _ projective with respect to @xmath59 in @xmath15_. 3 .
@xmath73 preserves morphisms in @xmath59 .
[ deflpet ] let @xmath2 be a concrete category such that its forgetful functor preserves epis and monos , @xmath1 a monad on @xmath2 , @xmath378 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) and ( b@xmath3813 ) .
a _ local pseudoequational @xmath1theory on @xmath13 _ is a nonempty collection @xmath584 of @xmath1algebra morphisms in @xmath59 with domain @xmath87 and finite codomain such that : a. for every finite set @xmath100 and @xmath585 , @xmath102 , there exists @xmath586 such that @xmath388 factors through @xmath62 , @xmath102 .
b. for every @xmath587 with codomain @xmath172 and every @xmath1algebra morphism @xmath390 with domain @xmath172 we have that @xmath588 . c. for every @xmath586 and @xmath589 we have that @xmath590 where @xmath395 is the factorization of @xmath396 . given an @xmath13generated algebra @xmath399 , we say that @xmath172 _ satisfies _ @xmath584 , denoted as @xmath591 , if every @xmath401 factors through some morphism in @xmath584 .
we denote by @xmath592 the _ finite @xmath13generated models of _ @xmath584 , that is : @xmath593 a class @xmath95 of finite @xmath13generated @xmath1algebras is _ defined _ by @xmath584 if @xmath594 .
let @xmath2 be a complete concrete category , @xmath1 a monad on @xmath2 , @xmath61 a factorization system on @xmath2 and @xmath378 a finite object .
a class @xmath95 of finite @xmath13generated algebras in @xmath15 is called a _ local pseudovariety of @xmath13generated @xmath1algebras _ if it is closed under @xmath59quotients , @xmath13generated @xmath60subalgebras and finite subdirect products .
[ lreitthm ] let @xmath2 be a concrete complete category such that its forgetful functor preserves epis , monos and products , @xmath1 a monad on @xmath2 , @xmath378 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) to ( b@xmath3813 ) .
then a class @xmath95 of finite @xmath13generated @xmath1algebras is a local pseudovariety of @xmath13generated @xmath1algebras if and only if is defined by a local pseudoequational @xmath1theory on @xmath13 .
additionally , local pseudovarieties of @xmath13generated @xmath1algebras are in one to
one correspondence with local pseudoequational @xmath1theories on @xmath13 .
[ deflpcet ] let @xmath3 be a concrete category such that its forgetful functor preserves monos , @xmath4 a comonad on @xmath3 , @xmath575 and @xmath61 a factorization system on @xmath3 .
assume ( b@xmath3811 ) and that @xmath80 preserves the morphisms in @xmath60 .
a _ local pseudocoequational @xmath4theory on @xmath28 _ is a nonempty collection @xmath595 of @xmath4coalgebra morphisms in @xmath60 with codomain @xmath596 and finite codomain such that : a. for every finite set @xmath100 and @xmath597 , @xmath102 , there exists @xmath598 such that @xmath388 factors through @xmath62 , @xmath102 .
b. for every @xmath599 with domain @xmath172 and every @xmath4coalgebra morphism @xmath451 with codomain @xmath172 we have that @xmath600 . c. for every @xmath598 and @xmath601 we have that @xmath602 where @xmath456 is the factorization of @xmath457 . with the previous definition , theorem [ lreitthm ] and duality , we have the following . [ leilpvar ]
let @xmath2 be a complete category , @xmath1 a monad on @xmath2 , @xmath61 a factorization system on @xmath2 and @xmath378 .
assume ( b@xmath3811 ) to ( b@xmath3813 ) .
let @xmath3 be a category that is dual to @xmath2 , @xmath28 the corresponding dual object of @xmath13 and let @xmath4 be the comonad on @xmath3 that is dual to @xmath1 which is defined as in proposition [ monadtocomonad ] .
then there is a one to
one correspondence between local varieties of @xmath13generated @xmath1algebras and local coequational @xmath4theories on @xmath28 .
[ exlpvv ] by fixing an object @xmath75 , we can get corresponding local versions of the eilenberg type correspondences showed in subsection [ secreiteil ] for pseudovarieties of @xmath13generated @xmath1algebras .
for example , the local version of ( 1 ) in example [ expvv1 ] reads as follows : there is a one to
one correspondence between pseudovarieties of @xmath13generated monoids and boolean algebras @xmath482 that are subalgebras of the complete atomic boolean algebra @xmath209 such that : a. every element in @xmath11 is a recognizable language on @xmath13
. b. @xmath11 is closed under left and right derivatives .
that is , for every @xmath580 and @xmath132 , @xmath581 . c. @xmath11 is closed under morphic preimages .
that is , for every homomorphism of monoids @xmath582 and @xmath580 , we have that @xmath583 .
we proved that eilenberg type correspondences = birkhoff s theorem for ( finite ) algebras + duality .
the main contribution of the present paper was to realize that the concept of a `` variety of languages '' that is used in eilenberg type correspondences corresponds to the dual of the ( pseudo)equational theory that defines the ( pseudo)variety of algebras , which was conjectured by the author in @xcite .
this not only allows us to understand where `` varieties of languages '' come from but also to get an abstract and general result that encompasses both existing and new eilenberg type correspondences .
our algebras of interest are @xmath1algebras , where @xmath1 is a monad on a category @xmath2 .
we stated and proved , under mild assumptions , categorical versions of birkhoff s theorem for @xmath1algebras and birkhoff s theorem for finite @xmath1algebras , the latter also known as reiterman s theorem for @xmath1algebras . in order to get eilenberg type correspondences we stated birkhoff s theorem for ( finite ) @xmath1algebras as a one to one correspondence between ( pseudo)varieties of @xmath1algebras and ( pseudo)equational @xmath1theories .
the previous observation led us to define the notion of a ( pseudo)equational @xmath1theory , i.e. , collections of `` ( pseudo)equations '' that are deductively closed .
the proof of birkhoff s theorem for @xmath1algebras is obtained from @xcite . on the other hand , the proof of birkhoff s theorem for finite @xmath1algebras follows the general idea used by the author in @xcite to prove a general eilenberg type correspondence and the idea that pseudovarieties of algebras are exactly directed unions of equational classes of finite algebras ( * ? ? ?
* proposition 4 ) .
it is worth mentioning that the proof of birkhoff s theorem for finite algebras in the present paper has the advantage of avoiding topological and profinite techniques , which are usually used to prove this theorem .
this help us to understand the proof of birkhoff s theorem for finite algebras in a more basic setting .
once we stated our categorical versions of birkhoff s theorem and birkhoff s theorem for finite algebras , with the use of duality , we stated our abstract eilenberg type correspondence theorems for varieties and pseudovarieties of @xmath1algebras .
then , in a similar way , in section [ seclocal ] , we derived corresponding local versions of birkhoff s theorem for ( finite ) @xmath13generated @xmath1algebras and their corresponding eilenberg type correspondences .
these local versions can be seen as particular instances of the previous work in which the category @xmath74 has only one object , say @xmath13 .
proofs are made in a similar way by restricting the kind of algebras to @xmath13generated @xmath1algebras .
thus the closure properties for local ( pseudo)varieties are closure under @xmath59quotients , @xmath13generated @xmath60subalgebras and ( finite ) subdirect products . from our abstract eilenberg
type correspondence theorems we derived both existing and new eilenberg type correspondences , including the ones in the following table : & pseudovariety version & local pseudovariety version & variety version & local variety version + semigroups & ( * ? ? ?
34s ) ( ex .
[ expvv1 ] ( 2 ) ) & @xcite ( ex . [ exlpvv ] ) & ex .
[ exvv1 ] ( 1 ) & ex .
[ exlvv ] + ordered semigroups & @xcite ( ex . [ expvv3 ] ) & @xcite ( ex . [ exlpvv ] ) & ex .
[ exvv3 ] & ex .
[ exlvv ] + monoids & ( * ? ? ?
34 ) ( ex .
[ expvv1 ] ( 1 ) ) & @xcite ( ex . [ exlpvv ] ) & ( * ? ? ?
39 ) ( ex . [ exjan11 ] ) & ex .
[ exlvv ] + ordered monoids & @xcite ( ex . [ expvv3 ] ) & @xcite ( ex . [ exlpvv ] ) & ex .
[ exvv3 ] & ex .
[ exlvv ] + groups & ex .
[ expvv1 ] ( 3 ) & ex .
[ exlpvv ] & ex .
[ exvv1 ] ( 2 ) & ex .
[ exlvv ] + ordered groups & ex . [ expvv3 ] & ex .
[ exlpvv ] & ex .
[ exvv3 ] & ex .
[ exlvv ] + monoid actions ( dynamical systems ) & ex . [ expvv1 ]
( 4 ) & ex .
[ exlpvv ] & ex .
[ exvv1 ] ( 3 ) & ex .
[ exlvv ] + semigroups with infinite exponentiation & cf .
@xcite ( ex .
[ expvv1 ] ( 5 ) ) & cf .
6.3 ) ( ex . [ exlpvv ] ) & ex .
[ exvv1 ] ( 4 ) & ex .
[ exlvv ] + @xmath7algebras for a finite field @xmath7 & @xcite ( ex . [ expvv4 ] ) & @xcite for @xmath603 ( ex . [ exlpvv ] ) & ex .
[ exvv4 ] & ex .
[ exlvv ] + idempotent semirings & @xcite ( ex . [ expvv5 ] ) & @xcite ( ex . [ exlpvv ] ) & ex .
[ exvv5 ] & ex .
[ exlvv ] + algebras of type @xmath91 in a variety & ex .
[ exuaf ] & ex .
[ exlpvv ] & ex .
[ exua ] & ex .
[ exlvv ] + ordered algebras of type @xmath91 in a variety & ex .
[ exuoaf ] & ex .
[ exlpvv ] & ex .
[ exuoa ] & ex .
[ exlvv ] + we will discuss some of the related work of this paper .
birkhoff s theorem : we discuss categorical approaches for birkhoff s theorem such as @xcite .
the main purpose of the present paper was to prove abstract eilenberg type correspondences which , in the case of varieties of @xmath1algebras , led us to state a birkhoff s theorem for @xmath1algebras in which every variety is defined by a unique collection of equations and vice versa .
this version is obtained from @xcite after defining the right notion of an equational @xmath1theory . in @xcite , the defining properties for a variety
are taken with respect to the factorization system @xmath61 where @xmath89 regular epi and @xmath90 mono , ( * ? ? ?
* definition 2.1 and 2.2 ) . in approaches such as @xcite the morphisms that represent equations are epimorphisms with projective domain , which is stated in condition ( b2 ) , and there is also the requirement of having enough projectives , which is implied by conditions ( b2 ) and ( b3 ) . in @xcite
, they work also with @xmath1algebras and a factorization system @xmath61 on @xmath15 . in the present paper ,
the factorization system @xmath61 is on @xmath2 which , under conditions ( b1 ) and ( b4 ) , is lifted to @xmath15 , but not every factorization system on @xmath15 is induced by one on the base category @xmath2 ( e.g. , on @xmath48 there is no factorization system that corresponds in any way to epimorphisms in the categories of monoids or of rings ) . in @xcite , the defining properties of a variety are closure under @xmath382split quotients , @xmath60subalgebras and products , where @xmath604 is the forgetful functor .
birkhoff s theorem for finite algebras : birkhoff s theorem for finite algebras , which is also known as reiterman s theorem @xcite , has been generalized in @xcite .
the approach in @xcite was to consider implicit operations .
implicit operations generalize the notion of terms .
equations given by implicit operations are the kind of equations that define pseudovarieties . the proof given in @xcite involves the use of topology in which the set of @xmath375ary implicit operations is the completion of the set of @xmath375ary terms . in @xcite , a topological approach is also considered by using uniformities , and it is also shown that pseudovarieties are exactly directed unions of equational classes of finite algebras ( * ? ? ?
* proposition 4 ) . in @xcite , a categorical approach
is considered to prove a reiterman s theorem for @xmath1algebras , this is done by using profinite techniques to define the notion of profinite equation which are the kind of equations that allow to define and characterize pseudovarieties . in @xcite , for a given monad @xmath1 on @xmath2 they define the profinite monad @xmath605 on the profinite completion @xmath606 of the category @xmath607 which is done by using limits ( in fact , right kan extensions ) .
the approach in the present paper do not use topological nor profinite techniques , and it is based in the fact that pseudovarieties are exactly directed unions of equational classes of finite algebras , see ( * ? ? ?
* proposition 4 ) and @xcite .
nevertheless , profinite and topological techniques can be easily brought to the scene in the present paper if we identify the family of morphisms @xmath385 by its limit , where @xmath384 is a pseudoequational @xmath1theory .
this would have led us to deal with profinite completions and topological spaces , in particular , profinite monoids , stone spaces and stone duality .
we prefer to avoid this approach for the following reasons : a. make the present work more accessible to some readers . b. to present a different approach without using topology and profinite techniques . c. eilenberg
type correspondences deal with pseudocoequational theories rather than its dual , i.e. , pseudoequational theories .
categorical eilenberg type correspondences : there are some categorical approaches for eilenberg type correspondences in the literature such as @xcite . in @xcite
only pseudovarieties are considered , i.e. , all the algebras are finite , while in @xcite as well as in the present paper we can also consider varieties of algebras .
in this respect , eilenberg type correspondences such as ( * ? ? ?
* theorem 39 ) or the ones derived in examples of subsection [ seceilv ] can not be derived from @xcite .
the work in @xcite subsumes the work made in @xcite .
the main setting in @xcite is to consider predual categories , i.e. , categories that are dual on finite objects .
the main purpose of this preduality is to define pseudovarieties of algebras on one category and varieties of languages on the other one . in @xcite ,
no duality is involved .
the definition of varieties of languages given in @xcite is restricted in the sense that it is always a boolean algebra , which in our present paper reduces to consider @xmath383 , @xmath608 , @xmath459 and @xmath609 , where @xmath11 is a fixed set .
type correspondences such as @xcite can not be derived from @xcite . in @xcite ,
all the algebras considered have a monoid structure which restricts the kind of algebras one can consider , e.g. , a semigroup version of eilenberg s theorem ( * ? ? ?
* theorem 34s ) can not be derived from @xcite .
those previous two limitations in the kind of varieties of languages and the kind of algebras one can consider are overcome in @xcite as well as in the present paper by considering algebras for a monad @xmath1 , which is the main idea in @xcite . in @xcite , the preduality considered as its main setting is lifted , under mild assumptions , to a full duality between one of the categories and the profinite completion of the other one . from this profinite completion ,
the concept of profinite equations is defined which are the kind of equations that define pseudovarieties of algebras . on the other hand ,
the definition of `` varieties of languages '' given in @xcite depends on finding a `` unary representation '' which is a set of unary operations on a free algebra satisfying certain properties ( * ? ? ?
* definition 37 ) and requires non trivial work . from this unary representation they construct syntactic algebras and define the kind of derivatives that define a variety of languages . in @xcite as well as in the present paper , the use of derivatives is not explicitely made which is captured in a more transparent and categorical way by using coalgebras , from this , the righ notion of derivatives easily follows by using duality and the defining properties of a @xmath1algebra ( epi)morphism ( see , e.g. , example [ exjan11 ] ) . the coalgebraic approach
used in the present paper gave us the advantage to obtain what we called an abstract eilenberg type correspondence in which the concept of `` variety of languages '' is the one of being a ( pseudo)coequational @xmath4theory , whose definition does not depend on finding the right notion of derivatives , contrary to @xcite , and does not depend on the existence of syntactic algebras , contrary to @xcite .
also , requirements considered in @xcite such as existence of `` unary representations '' or the fact that @xmath2 and @xmath3 are dual on finite objects are not needed to state our abstract eilenberg type correspondences . nevertheless , in specific applications such as , e.g. , example [ exuaf ] , the fact that the functor @xmath610 , which is part of the duality between @xmath48 and @xmath49 , preserves finite objects allowed us to identify the dual of the family @xmath385 with the boolean algebra @xmath437 . but
again , the fact that @xmath3 and @xmath2 are dual on finite objects does not play any role in our abstract eilenberg type correspondence .
the work in the present paper subsumes the work made by the author in @xcite from which most of the ideas presented in this paper were obtained .
the main idea of considering `` varieties of languages '' as coequations was initially made in ( * ? ? ?
* proposition 12 ) which was the starting point to suspect that `` varieties of languages '' are exactly duals of equational theories . with this in mind , the main work was focused on finding a proper definition of a ( pseudo)equational theory and state categorical versions of birkhoff s theorem and birkhoff s theorem for finite @xmath1algebras to get one to one correspondences between ( pseudo)varieties of @xmath1algebras and ( pseudo)equational @xmath1theories . as a consequence ,
we now clearly understand where `` varieties of languages '' come from and how to derive and find their defining properties in each particular case , e.g. , derivatives come from the properties that characterize a @xmath1algebra ( epi)morphism and closure under morphic preimages , which is a property that it is always present , comes from the substitution property in a ( pseudo)equational theory . the use of duality : related work such as @xcite have influenced and motivated the use of duality in language theory to characterize recognizable languages and to derive local versions of eilenberg type correspondences .
in fact , in this paper we show that duality is an ingredient to obtain ( abstract ) eilenberg type correspondences .
the most important aspect of this is in each concrete case of an eilenberg type correspondence , in which , by using the interaction between algebra and coalgebra and equations and coequations , one can easily find the right notion of derivatives as shown in the examples .
previous categorical approaches for eilenberg type correspondences such as @xcite have used duality , either explicitely or implicitely , but the fact that the dual of a pseudovariety of languages is exactly the pseudoequational theory that defines the given pseudovariety of algebras has not been brought to light to derive and understand eilenberg type correspondences .
for instance : a. none of the other categorical approaches to derive eilenberg
type correspondences relates the dual of a ( pseudo)variety of languages with a ( pseudo)equational theory or equations to obtain the defining properties of a ( pseudo)variety of languages .
b. derivatives are not directly obtained via duality in @xcite , no interaction between algebra and coalgebra or equations and coequations . c. approaches such as @xcite require that the categories @xmath56 and @xmath607 are dual in order to obtain an eilenberg type correspondence , which in our abstract eilenberg type correspondence theorem is not necessary .
it s is worth mentioning that some of the aspects of this paper have been previously studied , either implicitely or explicitely , in relation with eilenberg type correspondences .
in fact : a. in ( * ? ? ? * section 4 ) , it is mentioned , for the concrete case of monoids , that the dual of a `` local variety of languages '' will induce a set of ( in)equations . in the present paper , this is seen in example [ exlpvv ] for the free ( ordered ) monoid monad on @xmath48 ( @xmath50 ) .
b. in @xcite , recognizable subsets of an algebra with a single binary operation are studied , which in the present paper is the case of the local version of example [ exuaf ] for the type of algebras @xmath611 which consists of a single binary operation .
in fact , `` closure under residuals w.r.t .
singleton denominators '' in @xcite is the same as closure under derivatives with respect to @xmath611 in the sense of example [ exuaf ] . c. the duality between equations and coequations in the context of an eilenberg type correspondence
is studied for the first time in @xcite .
there , the `` varieties of languages '' considered have a closure property which is defined in terms of coequations ( * ? ? ?
* definition 40 ) .
d. the treatment of `` varieties of languages '' as sets of coequations was considered in ( * ? ? ?
* proposition 2 ) . there
, in the conclusions , was also conjectured that `` varieties of languages '' are exactly ( pseudo)coequational theories and their dual are the defining ( pseudo)equational theories for the ( pseudo)varieties of algebras .
syntactic algebras : another important observation and conclusion of the present paper is about the use of syntactic algebras . in eilenberg s original proof ( * ? ? ?
* theorem 34 ) the use of syntactic monoids ( semigroups ) ( * ? ? ?
* vii.1 ) helped to prove his theorem .
as in eilenberg s proof , the use of syntactic algebras was also used in @xcite for establishing eilenberg type correspondences .
categorical approaches such as @xcite generalized the concept of syntactic algebra . in @xcite
syntactic algebras are obtained , under mild assumptions , by means of a congruence , while in @xcite are obtained by using generalized pushouts , under the condition that @xmath73 preserves weak generalized pushouts . as we saw in the present paper ,
the use of syntactic algebras is not needed in order to establish abstract eilenberg type correspondences .
nonetheless , the study of syntactic algebras has their own importance in language theory and categorical generalizations of them such as @xcite might deserve a further study .
some of the future work that can be done based on the work presented in this paper include the following : a. the relation of equational @xmath1theories with monad morphisms @xmath612 as in @xcite .
b. to find new eilenberg type correspondences as an application of our abstract ( local ) eilenberg type theorem for ( pseudo)varieties of @xmath1algebras .
c. to study eilenberg
type correspondences for other classes of algebras , e.g. , eilenberg type correspondences for quasivarieties .
which can be made by modifying the notion of an equational theory given in the present paper by allowing @xmath1algebra morphisms with arbitrary domain ( not necessarily a free one ) see , e.g. , @xcite .
d. to find applications of the eilenberg type correspondences we can derive from the present paper .
one example of this could be to characterize the ( pseudo)equational @xmath4theory that defines a particular ( pseudo)variety of @xmath1algebras .
this kind of problem has been studied before in which the pseudovariety of aperiodic monoids is defined by the variety of languages in which every language is star
free @xcite .
e. the study and applications of the dual theorems in this paper .
that is , cobirkhoff s theorem @xcite , coreiterman s theorem and a new subject that we can call coeilenberg
type correspondences which are naturally defined as one to
one correspondences between ( pseudo)covarieties of @xmath4coalgebras and ( pseudo)equational @xmath1theories . f. to develop and study a general theory for syntactic algebras .
as we mentioned , syntactic algebras are not used in the present paper to establish abstract eilenberg type correspondences . in a previous research made by the author in @xcite , to prove a general eilenberg type theorem , syntactic algebras were also considered and constructed abstractly as a generalized pushout ( * ? ? ?
* proposition 10 ) .
general syntactic algebras were also considered and constructed in @xcite .
i would like to thank marcello bonsangue , jan rutten and alexander kurz for their support , comments , and suggestions during the writing of this paper .
i thank henning urbat , ji admek , liang ting chen and stefan milius for earlier discussions we had on this subject .
i also thank anonymous referees for their valuable comments and suggestions on an earlier version of this paper .
admek , j. , s. milius , r. myers , and h. urbat , _ generalized eilenberg theorem i : local varieties of lamguages , _ foundations of software science and computation structures : 17th international conference , fossacs 2014 , 366380 .
\(a ) at ( 0,0 ) @xmath66 ; ( b ) at ( 2,0 ) @xmath66 ; ( c ) at ( 0,1.5 ) @xmath66 ; ( d ) at ( 2,1.5 ) @xmath66 ; ( a ) edge [ ] node [ below ] @xmath67 ( b ) ; ( c ) edge [ ] node [ above ] @xmath615 ( d ) ; ( c ) edge [ ] node [ left ] @xmath616 ( a ) ( d ) edge [ ] node [ right ] @xmath617 ( b ) ; using the diagonal fill in we get a morphism @xmath70 such that @xmath618 .
now , from @xmath619 , by using the fact that @xmath67 is mono , we have @xmath620 , i.e. , @xmath67 is iso .
therefore @xmath621 since @xmath64 and @xmath67 is iso .
_ let @xmath2 be a category , @xmath14 a monad on @xmath2 , and @xmath61 a proper factorization system on @xmath2 .
if @xmath73 preserves the morphisms in @xmath59 then @xmath15 inherits the same @xmath61 factorization system . that is : _
a. given @xmath122 and @xmath622 such that @xmath623 , if @xmath624 is factored as @xmath63 in @xmath2 where @xmath625 , @xmath65 and @xmath626 is the domain of @xmath67 , then there exists a unique @xmath627 such that @xmath628 and @xmath67 and @xmath68 are @xmath1algebra morphisms .
b. given any commutative diagram + \(a ) at ( 0,0 ) @xmath629 ; ( b ) at ( 2,0 ) @xmath630 ; ( c ) at ( 0,1.5 ) @xmath81 ; ( d ) at ( 2,1.5 ) @xmath80 ; ( a ) edge [ ] node [ below ] @xmath67 ( b ) ; ( c ) edge [ ] node [ above ] @xmath68 ( d ) ; ( c ) edge [ ] node [ left ] @xmath62 ( a ) ( d ) edge [ ] node [ right ] @xmath69 ( b ) ; + in @xmath15 with @xmath64 and @xmath65 , the unique _ diagonal fill in _ morphism @xmath70 such that the diagram + \(a ) at ( 0,0 ) @xmath629 ; ( b ) at ( 2,0 ) @xmath630 ; ( c ) at ( 0,1.5 ) @xmath81 ; ( d ) at ( 2,1.5 ) @xmath80 ; ( a ) edge [ ] node [ below ] @xmath67 ( b ) ; ( c ) edge [ ] node [ above ] @xmath68 ( d ) ; ( c ) edge [ ] node [ left ] @xmath62 ( a ) ( d ) edge [ ] node [ above ] @xmath70 ( a ) ( d ) edge [ ] node [ right ] @xmath69 ( b ) ; + commutes is a morphism in @xmath15 .
a. as @xmath631 and @xmath63 , we get the following commutative diagram : + \(a ) at ( 0,0 ) @xmath81 ; ( b ) at ( 2,0 ) @xmath629 ; ( c ) at ( 4,0 ) @xmath80 ; ( d ) at ( 0,2 ) @xmath632 ; ( e ) at ( 2,2 ) @xmath633 ; ( f ) at ( 4,2 ) @xmath634 ; ( a ) edge [ ] node [ above ] @xmath68 ( b ) ( d ) edge [ ] node [ below ] @xmath635 ( e ) ; ( b ) edge [ ] node [ below ] @xmath67 ( c ) ; ( d ) edge [ ] node [ left ] @xmath636 ( a ) ( e ) edge [ ] node [ below ] @xmath637 ( f ) ( f ) edge [ ] node [ right ] @xmath638 ( c ) ; ( e ) edge [ ] node [ right ] @xmath639 ( b ) ; ( a ) edge [ bend right , looseness=0.7 ] node [ below ] @xmath62 ( c ) ( d ) edge [ bend left , looseness=0.7 ] node [ above ] @xmath640 ( f ) ; + where @xmath627 is obtained by the diagonal fill in property since @xmath641 .
now , we prove that @xmath642 .
in fact , a. we prove that @xmath643 .
in fact , we have the following commutative diagram : + \(a ) at ( 5.7,2 ) @xmath81 ; ( b ) at ( 2,0 ) @xmath629 ; ( c ) at ( 3.8,0 ) @xmath80 ; ( d ) at ( 2,4 ) @xmath632 ; ( e ) at ( 2,2 ) @xmath633 ; ( f ) at ( 3.8,2 ) @xmath634 ; ( g ) at ( 0,4 ) @xmath81 ; ( h ) at ( 0,2 ) @xmath629 ; ( i ) at ( 5.7,0 ) @xmath629 ; ( a ) edge [ ] node [ right ] @xmath68 ( i ) ( d ) edge [ ] node [ left ] @xmath635 ( e ) ( g ) edge [ ] node [ left ] @xmath68 ( h ) ; ( b ) edge [ ] node [ below ] @xmath67 ( c ) ; ( i ) edge [ ] node [ below ] @xmath67 ( c ) ; ( d ) edge [ ] node [ above ] @xmath636 ( a ) ( e ) edge [ ] node [ below ] @xmath637 ( f ) ( f ) edge [ ] node [ right ] @xmath638 ( c ) ( e ) edge [ ] node [ right ] @xmath639 ( b ) ( a ) edge [ ] node [ below ] @xmath62 ( c ) ( d ) edge [ ] node [ right ] @xmath640 ( f ) ( g ) edge [ ] node [ above ] @xmath644 ( d ) ( h ) edge [ ] node [ below ] @xmath645 ( e ) ; + from this , starting from @xmath81 at the top left corner and finishing at @xmath80 we have that @xmath646 . from that equation , using the fact that @xmath647 , since @xmath648 , and the fact that @xmath67 is mono and @xmath68 is epi , we have that @xmath643 .
b. we prove that @xmath649 .
in fact , we have the following commutative diagram : + \(a ) at ( 0,0 ) @xmath633 ; ( b ) at ( 3,0 ) @xmath629 ; ( c ) at ( 6,0 ) @xmath80 ; ( d ) at ( 3,1 ) @xmath634 ; ( e ) at ( -1,3 ) @xmath650 ; ( f ) at ( 1,3 ) @xmath632 ; ( g ) at ( 5,3 ) @xmath634 ; ( h ) at ( 7,3 ) @xmath629 ; ( i ) at ( 3,3 ) @xmath651 ; ( j ) at ( 3,5 ) @xmath632 ; ( k ) at ( 0,6 ) @xmath652 ; ( l ) at ( 3,6 ) @xmath650 ; ( m ) at ( 6,6 ) @xmath633 ; ( f ) edge [ ] node [ right ] @xmath635 ( a ) ( k ) edge [ ] node [ left ] @xmath653 ( e ) edge [ ] node [ above ] @xmath653 ( l ) ( j ) edge [ ] node [ above ] @xmath635 ( m ) ; ( b ) edge [ ] node [ below ] @xmath67 ( c ) ; ( h ) edge [ ] node [ right ] @xmath67 ( c ) ; ( a ) edge [ ] node [ below ] @xmath639 ( b ) edge [ ] node [ above ] @xmath637 ( d ) ( g ) edge [ ] node [ left ] @xmath638 ( c ) ( m ) edge [ ] node [ left ] @xmath637 ( g ) edge [ ] node [ right ] @xmath639 ( h ) ( d ) edge [ ] node [ above ] @xmath638 ( c ) ( l ) edge [ ] node [ above ] @xmath654 ( m ) ( k ) edge [ ] node [ above ] @xmath655 ( j ) edge [ ] node [ right ] @xmath656 ( f ) edge [ ] node [ right ] @xmath657 ( i ) ( e ) edge [ ] node [ left ] @xmath658 ( a ) ( f ) edge [ ] node [ above ] @xmath659 ( d ) ( i ) edge [ ] node [ right ] @xmath660 ( d ) edge [ ] node [ above ] @xmath661 ( g ) ( j ) edge [ ] node [ right ] @xmath640 ( g ) ; + then by following the external arrows we get that @xmath662 .
then , from that equation , since @xmath67 is mono and @xmath653 is epi , we have that @xmath649 .
b. put @xmath622 and @xmath663 .
then we have that @xmath664 . from that equality ,
since @xmath67 is mono and @xmath635 is epi , we get @xmath665 , i.e. , @xmath666
. the proof for birkhoff s theorem for @xmath1algebras can be made by following the same ideas for standard proofs of birkhoff s theorem , see , e.g. , @xcite
. in our case we deal with equational @xmath1theories and we have a fixed the subcategory @xmath74 of `` variables '' , which is the main difference with respect to some other versions such as @xcite
. we will derive theorem [ birkthm ] from the following basic facts and some facts from @xcite .
[ lemmafree ] let @xmath2 be a category , @xmath61 a factorization system on @xmath2 , @xmath14 a monad on @xmath2 and @xmath74 a full subcategory of @xmath2 .
assume ( b2 ) .
let @xmath104 be an equational @xmath1theory on @xmath74 .
then @xmath667 for every @xmath75 .
\(b ) at ( 4,2 ) @xmath110 ; ( c ) at ( 0,2 ) @xmath108 ; ( d ) at ( 2,2.4 ) @xmath109 ; ( e ) at ( 6,2 ) @xmath82 ; ( c ) edge [ ] node [ above ] @xmath111 ( d ) ; ( e ) edge [ ] node [ above ] @xmath112 ( b ) ; ( d ) edge [ ] node [ above ] @xmath114 ( b ) ; ( c ) edge [ bend right , looseness=0.7 ] node [ below ] @xmath62 ( b ) ; ( c ) edge [ bend left , looseness=0.7 ] node [ above ] @xmath69 ( e ) ; where @xmath669 is obtained from @xmath62 and @xmath112 using assumption ( b2 ) and @xmath114 is obtained from the fact that @xmath173 is an equational @xmath1theory . therefore , @xmath62 factors through @xmath111 and hence @xmath667 .
[ injth ] let @xmath2 be a category , @xmath61 a factorization system on @xmath2 , @xmath14 a monad on @xmath2 and @xmath74 a full subcategory of @xmath2 .
assume ( b1 ) and ( b2 ) .
for @xmath670 , let @xmath671 be an equational @xmath1theory on @xmath74 . if @xmath672 then @xmath673
as @xmath672 , there exists @xmath75 such that @xmath674 , i.e. , there is no isomorphism @xmath675 such that @xmath676 .
we have that @xmath677 or @xmath678 .
in fact , if we assume by contradiction that @xmath679 and @xmath680 then , from the fact that @xmath679 , we get the commutative diagram : \(b ) at ( 5,2 ) @xmath681 ; ( c ) at ( 0,2 ) @xmath82 ; ( d ) at ( 2.5,2 ) @xmath682 ; ( c ) edge [ ] node [ above ] @xmath683 ( d ) ; ( d ) edge [ ] node [ above ] @xmath684 ( b ) ; ( c ) edge [ bend left , looseness=0.7 ] node [ above ] @xmath685 ( b ) ; i.e. , there exists @xmath686 such that @xmath687 . similarly , from the fact that @xmath680 , we get that there exists @xmath688 such that @xmath689 .
hence we have that : @xmath690 which implies that @xmath691 since @xmath683 is epi by ( b1 ) .
similarly , @xmath692 , which implies that @xmath693 is an isomorphism such that @xmath689 which is a contradiction .
hence @xmath677 or @xmath678 and , by the previous lemma , @xmath694 , which implies that @xmath673 .
[ hspclosureb ] let @xmath2 be a complete category , @xmath14 a monad on @xmath2 , @xmath61 a factorization system on @xmath2 and @xmath74 a full subcategory of @xmath2 .
assume ( b1 ) , ( b2 ) and ( b4 ) .
let @xmath173 be an equational @xmath1theory on @xmath74 .
then @xmath180 is a variety of @xmath1algebras .
a. @xmath180 is closed under @xmath59quotients : let @xmath623 with @xmath695 and let @xmath78 .
let @xmath696 such that @xmath75 , then we have the following commutative diagram : + \(b ) at ( 4,2 ) @xmath81 ; ( c ) at ( 0,2 ) @xmath82 ; ( d ) at ( 2,2 ) @xmath110 ; ( e ) at ( 6,2 ) @xmath80 ; ( c ) edge [ ] node [ above ] @xmath112 ( d ) ( b ) edge [ ] node [ above ] @xmath68 ( e ) ; ( d ) edge [ ] node [ above ] @xmath697 ( b ) ; ( c ) edge [ bend left , looseness=0.6 ] node [ above ] @xmath698 ( b ) ; ( c ) edge [ bend left , looseness=0.7 ] node [ above ] @xmath62 ( e ) ; + where @xmath699 was obtained from @xmath62 using ( b2 ) and @xmath700 from the fact that @xmath695 . therefore @xmath62 factors through @xmath112 , i.e. , @xmath701 . b. @xmath180 is closed under @xmath60subalgebras : let @xmath623 with @xmath695 and let @xmath702 .
let @xmath696 such that @xmath703 , then we have the following commutative diagram : + \(b ) at ( 4,2 ) @xmath80 ; ( c ) at ( 0,2 ) @xmath82 ; ( d ) at ( 2,2.4 ) @xmath110 ; ( e ) at ( 6,2 ) @xmath81 ; ( c ) edge [ ] node [ above ] @xmath112 ( d ) ; ( b ) edge [ ] node [ above ] @xmath67 ( e ) ; ( d ) edge [ ] node [ above ] @xmath698 ( b ) ; ( c ) edge [ bend right , looseness=0.7 ] node [ below ] @xmath62 ( b ) ; ( d ) edge [ bend left , looseness=0.7 ] node [ above ] @xmath704 ( e ) ; + where @xmath705 was obtained from the fact that @xmath695 , and @xmath706 was obtained by the diagonal fill in property of the factorization system @xmath61 . since @xmath67 is mono , from @xmath707 , we get @xmath708 which implies that @xmath701 .
c. @xmath180 is closed under products : let @xmath709 , @xmath102 , and let @xmath710 be their product in @xmath15 with projections @xmath476 .
let @xmath175 such that @xmath75 , then we have the following commutative diagram : + \(b ) at ( 4,2 ) @xmath81 ; ( c ) at ( 0,2 ) @xmath82 ; ( d ) at ( 2,2.4 ) @xmath110 ; ( e ) at ( 6,2 ) @xmath711 ; ( c ) edge [ ] node [ above ] @xmath112 ( d ) ; ( b ) edge [ ] node [ above ] @xmath712 ( e ) ; ( d ) edge [ ] node [ above ] @xmath69 ( b ) ; ( c ) edge [ bend right , looseness=0.7 ] node [ below ] @xmath62 ( b ) ; ( d ) edge [ bend left , looseness=0.7 ] node [ above ] @xmath713 ( e ) ; + where @xmath714 was obtained from the fact that @xmath709 , and @xmath715 was obtained by the universal property of the product .
finally , we have that @xmath716 since @xmath717 for every @xmath102 .
theorem [ birkthm ] follows from @xcite as follows .
we have that the facts in @xcite hold for any @xmath718category as it is mentioned before ( * ? ? ?
* example 1 ) ( see @xcite for basic facts and examples about @xmath718categories ) .
now , by the assumptions of theorem [ birkthm ] we have that the category @xmath15 is an @xmath718category , by ( * ? ? ?
* corollary 15.21 ) , and hence a class @xmath95 of @xmath1algebras , viewed as a full subcategory of @xmath15 , is @xmath719equational in @xmath15 if and only if @xmath95 is a variety . in this case
, the inclusion functor @xmath720 has a left adjoint @xmath721 such that its unit @xmath722 is such that all its componets are in @xmath723 and , by the assumptions , the equational @xmath1theory @xmath724 defines @xmath95 ( see , e.g. , ( * ? ? ?
* proposition 3 and remark 1 ) ) .
the fact that @xmath173 is an equational @xmath1theory follows from naturality of @xmath725 .
finally , by proposition [ hspclosureb ] every equational @xmath1theory defines a variety and the correspondence between equational @xmath1theories and varieties is bijective by proposition [ injth ] and uniqueness of left adjoint .
a. for every @xmath728 we have that @xmath729 ; b. if @xmath730 , where @xmath731 a monoid congruence of @xmath184 , @xmath102 , then we have that @xmath732 ; c. for every @xmath216 , if @xmath733 and @xmath734 denotes the quotient morphism , then for each monoid morphism @xmath735 we have @xmath736 .
a. @xmath738 and it is a subalgebra of @xmath739 .
b. @xmath740 is closed under left and right derivatives .
that is , if @xmath741 and @xmath132 then @xmath742 , where @xmath213 and @xmath214 , @xmath215 . c. @xmath737 is closed under morphic preimages
that is , for every @xmath216 , homomorphism of monoids @xmath217 and @xmath741 , we have that @xmath743 . by ( * ? ? ?
* corollary 8) we have that the monoid @xmath746 is the syntactic monoid of @xmath284 .
the universal property of the syntactic monoid of @xmath284 is , by duality , the property that @xmath745 .
this property of @xmath525 being the dual of the syntactic monoid of @xmath284 was also mentioned in ( * ? ? ?
* section 6 ) .
( @xmath749 ) : consider the canonical epimorphism of monoids @xmath750 , @xmath751 . then , by duality , i.e. , applying @xmath752 , gives us the monomorphism @xmath753 which implies that @xmath754 since @xmath755 .
lemma [ lj2 ] says that @xmath765 for every @xmath207 , since @xmath766 for every monoid congruence @xmath629 of @xmath184 ( * ? ? ?
* proposition 15 ) .
lemma [ lj1 ] together with property i ) of @xmath726 being a variety of languages imply that @xmath767 is closed under left and right derivatives .
that is , every variety of languages @xmath726 satisfies properties i ) and ii ) of a coequational @xmath4theory .
now we show that @xmath726 also satisfies property iii ) of a coequational @xmath4theory .
( @xmath748 ) : assume @xmath776 , i.e. , there exists @xmath777 with @xmath778 such that @xmath779 .
now , we have @xmath780 where the last implication follows from lemma [ lj3 ] since @xmath778 . finally , from @xmath778 we get @xmath781 .
this finishes the proof of the claim .
+ from the claim we have that @xmath772 .
a. properties i ) and ii ) of @xmath737 being a coequational @xmath4theory say that @xmath740 is a @xmath4subcoalgebra of @xmath209 .
in particular , for every @xmath741 we have @xmath782 .
b. to prove property ii ) we show that for a monoid congruence @xmath731 of @xmath184 , @xmath102 , the @xmath4coalgebra @xmath783 is the @xmath4subcoalgebra of @xmath209 generated by the family @xmath784 .
we show this by duality , i.e. , in the category of monoids .
we have the following setting : + \(a ) at ( 3,2 ) @xmath184 ; ( b ) at ( 6,0 ) @xmath785 ; ( c ) at ( 3,0 ) @xmath786 ; ( d ) at ( 0,0 ) @xmath787 ; ( b ) edge [ ] node [ below ] @xmath788 ( c ) ; ( a ) edge [ ] node [ above ] @xmath789 ( d ) edge [ ] node [ above ] @xmath790 ( b ) ; ( a ) edge [ ] node [ left ] @xmath725 ( c ) ( c ) edge [ ] node [ below ] @xmath563 ( d ) ; + where : 1 .
@xmath791 is the canonical homomorphism , @xmath565 , 2 .
@xmath786 is the product @xmath792 with projections @xmath793 , @xmath565 , 3 .
@xmath725 is obtained from @xmath789 , @xmath565 , by the universal property of @xmath786 , and 4 .
@xmath794 is the factorization of @xmath725 , i.e. , @xmath795 .
+ now we prove , by duality , that the @xmath4coalgebra @xmath783 is the least @xmath4subcoalgebra of @xmath209 containing each of @xmath796 .
let @xmath797 be an epimorphism of monoids such that each @xmath789 factors through @xmath68 , @xmath565 .
that is , there exists @xmath798 such that @xmath799 , @xmath565 .
therefore , @xmath800 , @xmath565 , and hence @xmath801 , which means that there exists @xmath802 such that @xmath803 .
+ now , @xmath737 satisfying property ii ) of a variety of languages follows from the observation above .
in fact , if @xmath740 contains @xmath796 , @xmath102 , then , by using the fact that @xmath740 is a @xmath4subcoalgebra of @xmath209 , it contains the least @xmath4subcoalgebra of @xmath209 containing each of @xmath796 , @xmath102 , which is @xmath783 . c. let @xmath216 , @xmath804 and @xmath734 be the quotient morphism .
let @xmath735 be a monoid morphism .
we have to show that @xmath805 .
in fact , let @xmath806 , i.e. , @xmath807 is of the form @xmath808 for some @xmath809 . define @xmath810 as @xmath811 .
then we have that @xmath812 which by i ) implies that @xmath813 , since @xmath814 . since @xmath737 is a coequational @xmath4theory then @xmath815 .
to finish the proof we prove the following : + _ claim : _ @xmath816 .
+ let @xmath777 , then : + ( @xmath749 ) : assume that @xmath817
. then there exists @xmath818 such that @xmath819 .
this implies that @xmath820 with @xmath818 , i.e. , @xmath821 which means that @xmath822 .
+ ( @xmath748 ) : assume that @xmath822 , i.e. , @xmath821
. then there exists @xmath818 such that @xmath823 .
this implies that @xmath819 with @xmath818 , i.e. , @xmath817 .
the duality between @xmath50 and @xmath51 is given by the hom - set functors @xmath824 and @xmath825 , where @xmath826 is the two element chain ` schizophrenic ' object in @xmath50 and in @xmath51 .
note that for any @xmath827 the object @xmath828 is the set of downsets of @xmath829 with the inclusion order and for any @xmath830 the object @xmath831 is the set of all completely join prime elements of @xmath172 with the order inherited from @xmath172 .
remember that an element @xmath832 of @xmath172 is _ completely join prime _ if @xmath833 implies @xmath834 for some @xmath488 .
a variety of ordered algebras in @xmath95 is defined by an equational @xmath221theory @xmath835 which by duality gives us the coequational @xmath4theory @xmath836 which is equivalently defined by the image of every embedding @xmath837 , i.e. , we define @xmath838 .
closure of @xmath210 under derivatives with respect to the type @xmath91 follows from the fact that each morphism @xmath112 in an equational @xmath126theory is a homomorphism of ordered algebras .
in fact , similar to example [ exua ] b ) , we have that for every for every @xmath226 of arity @xmath227 , every @xmath228 , every @xmath229 , @xmath230 , @xmath234 and @xmath839 we have @xmath840 , where the function @xmath245 is defined for every @xmath246 as @xmath247 .
we only need to prove that @xmath841 .
in fact , for any @xmath842 in @xmath110 we have that @xmath843 , where @xmath285 and @xmath844 are in the @xmath532th position , which implies that @xmath845 since @xmath839 .
therefore , @xmath210 is closed under derivatives with respect to the type @xmath91 .
+ conversely , any @xmath846 closed under derivatives with respect to the type @xmath91 such that @xmath11 is a subalgebra of @xmath847 will define , by duality , the surjective function @xmath848 such that @xmath849 , @xmath850 and @xmath580 , which is a morphism in @xmath851 .
we only need to show that for every @xmath219 , @xmath229 , @xmath852 and @xmath853 the inequality @xmath854 implies that @xmath855 ( see ( * ? ? ?
* proposition ) ) .
in fact , assume that @xmath854 , i.e. , for every @xmath580 we have that @xmath856 .
now , assume by contradiction that @xmath857 , i.e. , there exists @xmath858 such that @xmath859 and @xmath860 , i.e. , @xmath861 and @xmath862 with @xmath863 by closure under derivatives with respect to the type @xmath91 , which contradics the fact that @xmath854 .
therefore @xmath271 is a @xmath221algebra morphism in @xmath59 .
a variety of @xmath7algebras is defined by an equational @xmath1theory @xmath866 which by duality gives us the coequational @xmath4theory @xmath867 which is equivalently defined by the image of every monomorphism @xmath868 , i.e. , we define @xmath869 .
closure of @xmath210 under left and right derivatives follows from the fact that each morphism @xmath112 in an equational @xmath1theory is a homomorphism of @xmath7algebras .
in fact , for every @xmath870 and @xmath871 we have that @xmath872 where the function @xmath873 is defined as @xmath874 , where @xmath66 is the product operation in @xmath875 , @xmath246 .
note that @xmath876 since for any @xmath877 and @xmath878 we have that @xmath879 since @xmath871 .
therefore , @xmath880 , i.e. , @xmath210 is closed under right derivatives .
closure under left derivatives is proved in a similar way .
+ conversely , any @xmath881 closed under left and right derivatives such that @xmath11 is a subspace of @xmath882 will define , by duality , the surjective function @xmath883 such that @xmath849 , @xmath353 and @xmath580 , which is a morphism in @xmath884 .
we only need to show that for every @xmath885 the equality @xmath886 implies that @xmath887 and @xmath888 .
in fact , assume that @xmath886 , i.e. , for every @xmath580 we have that @xmath889 .
now , assume by contradiction that @xmath890 , i.e. , there exists @xmath858 such that @xmath891 , i.e. , @xmath892 with @xmath893 by closure under right derivatives , which is a contradiction .
the equality @xmath888 is proved in a similar way by using closure under left derivatives . therefore @xmath271 is a @xmath1algebra morphism in @xmath59 . and @xmath903 is defined as @xmath904 which is well defined by property iii )
. we should use a notation like @xmath905 for the relation defined above , but we will denote it by @xmath895 for simplicity . it will be clear from the context to which @xmath895 we are refering to in each case .
if @xmath906 then @xmath907 is defined as @xmath908 the unit of the monad is defined as @xmath909 and the multiplication as : @xmath910 where each @xmath911 is such that @xmath912 and @xmath913 , @xmath914 and @xmath915 .
+ we have that @xmath15 is the category of idempotent semirings .
by definition we have that @xmath919 now , every element in @xmath357 is of the form @xmath920 , which by property i ) and iii ) of the definition of @xmath895 we have that : @xmath921 therefore , by using the defining properties of @xmath895 we have that every element in @xmath922 is equivalent to a unique element of the form : @xmath923 where uniqueness follows since @xmath924 is the free join semilattice .
hence , the join semilattice homomorphism @xmath925 given by : @xmath926 is an isomorphism in @xmath54 .
we considered @xmath356 .
as every semiring is an @xmath59quotient of @xmath918 , by the previous lemma we have that condition ( b3 ) is satisfied .
+ now , in the definition of the operator @xmath206 we should formally have that @xmath210 is a subspace of @xmath927 but for simplicity we work with @xmath209 which is isomorphic to @xmath927 in @xmath55 under the correspondence @xmath928 and @xmath929 , @xmath930 and @xmath744 , where @xmath931 and @xmath932 are defined as @xmath933 and @xmath934 . + the duality between @xmath54 and @xmath55 is given by the hom - set functors @xmath935 and @xmath936 , where @xmath937 is the two element join semilattice .
+ a variety of idempotent semirings is defined by an equational @xmath1theory @xmath938 which by duality gives us the coequational @xmath4theory @xmath939 which is equivalently defined by the image of every monomorphism @xmath940 , i.e. , we define @xmath941 .
closure of @xmath210 under left and right derivatives follows from the fact that each morphism @xmath112 in an equational @xmath1theory is a homomorphism of idempotent semirings .
in fact , for every @xmath942 and @xmath943 we have that @xmath944 where the function @xmath945 is defined as @xmath946 , where @xmath66 is the product operation in @xmath875 , @xmath246 .
note that @xmath947 since for any @xmath878 we have that @xmath948 since @xmath943 .
therefore , @xmath949 , i.e. , @xmath210 is closed under right derivatives .
closure under left derivatives is proved in a similar way .
+ conversely , any @xmath950 closed under left and right derivatives such that @xmath11 is a subspace of @xmath951 will define , by duality , the surjective function @xmath952 such that @xmath953 , @xmath215 and @xmath580 , which is a morphism in @xmath954 .
we only need to show that for every @xmath215 and @xmath955 the equality @xmath956 implies that @xmath957 and @xmath958 .
in fact , assume that @xmath956 , i.e. , for every @xmath580 we have that @xmath959 .
now , assume by contradiction that @xmath960 , i.e. , there exists @xmath858 such that @xmath961 , i.e. , @xmath962 with @xmath963 by closure under right derivatives , which is a contradiction .
the equality @xmath958 is proved in a similar way by using closure under left derivatives . therefore @xmath271 is a @xmath1 algebra morphism in @xmath59 .
[ hspclosurer ] let @xmath2 be a complete concrete category such that its forgetful functor preserves epis , monos and products , @xmath1 a monad on @xmath2 , @xmath74 a full subcategory of @xmath74 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) , ( b@xmath3812 ) and ( b@xmath3814 ) .
let @xmath384 be a pseudoequational @xmath1theory on @xmath74 .
then @xmath423 is a pseudovariety of @xmath1algebras .
a. @xmath423 is closed under @xmath59quotients : let @xmath623 with @xmath966 and let @xmath78 .
we have that @xmath80 is finite since @xmath68 is epi and the forgetful functor from @xmath2 to @xmath48 preserves epis .
let @xmath967 , @xmath75 . using ( b@xmath3812 )
, there exists @xmath699 such that @xmath968 . as @xmath966 then @xmath698 factors through some @xmath503 as @xmath969 . then @xmath970 with @xmath503 , i.e. , @xmath971 .
b. @xmath423 is closed under @xmath60subalgebras : let @xmath623 with @xmath966 and let @xmath702 .
we have that @xmath80 is finite since @xmath67 is mono and the forgetful functor from @xmath2 to @xmath48 preserves monos .
let @xmath967 , @xmath703 .
as @xmath966 then @xmath972 factors through some @xmath389 as @xmath973 .
let @xmath479 and @xmath974 be the factorizations of @xmath62 and @xmath69 , respectively .
we have that @xmath975 where @xmath976 and @xmath977 .
then by uniqueness of the factorization we have that there is an isomorphism @xmath978 such that @xmath979 .
therefore @xmath980 with @xmath389 , i.e. , @xmath971 .
c. @xmath423 is closed under finite products : let @xmath981 , @xmath102 with @xmath100 finite , and let @xmath710 be their product in @xmath15 with projections @xmath476 .
we have that @xmath81 is finite since the forgetful functor from @xmath2 to @xmath48 preserves products , @xmath100 is finite , and each @xmath711 is finite .
let @xmath178 , @xmath75 .
as @xmath981 then @xmath982 factors through some @xmath983 as @xmath984 . since @xmath384 is a pseudoequational @xmath1theory there exists @xmath389 such that every @xmath474 factors through @xmath68 as @xmath985 , @xmath102 .
let @xmath986 be the codomain of @xmath68 .
now , by definition of @xmath172 there exists @xmath987 such that @xmath988 . as @xmath989 for every @xmath102 , then @xmath990 , @xmath389 , which means that @xmath966 .
let @xmath2 be a complete concrete category such that its forgetful functor preserves epis , monos and products @xmath1 a monad on @xmath2 , @xmath74 a full subcategory of @xmath74 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) and ( b@xmath3814 ) .
let @xmath95 be a pseudovariety of @xmath1algebras .
then @xmath991 is a pseudoequational @xmath1theory on @xmath74 .
a. let @xmath75 , @xmath100 a finite set and @xmath993 , @xmath102 .
let @xmath101 be the codomain of @xmath388 .
let @xmath710 with projections @xmath994 .
we have @xmath97 .
now , by definition of @xmath172 , there exists @xmath178 such that @xmath995 .
let @xmath479 be the factorization of @xmath62 with @xmath996 and @xmath997 .
we have that @xmath998 .
then @xmath501 is a morphism in @xmath999 such that every @xmath388 factors through @xmath501 .
b. let @xmath75 , @xmath1000 with codomain @xmath97 , and @xmath1001 .
we have that @xmath1002 is finite and that @xmath405
. therefore @xmath1003 .
c. let @xmath105 , @xmath1004 with codomain @xmath97 , and @xmath393 . let @xmath395 be the factorization of @xmath396 such that @xmath1005 and @xmath1006 .
then @xmath998 , which implies @xmath1007 .
[ lemmafr ] let @xmath2 be a complete concrete category such that its forgetful functor preserves epis , monos and products , @xmath1 a monad on @xmath2 , @xmath74 a full subcategory of @xmath74 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) , ( b@xmath3812 ) and ( b@xmath3814 ) .
let @xmath384 be a pseudoequational @xmath1theory on @xmath74 .
let @xmath75 and @xmath389 with codomain @xmath399 , then @xmath966 .
\(b ) at ( 3,1.3 ) @xmath82 ; ( c ) at ( 3,0 ) @xmath1010 ; ( d ) at ( 6,2 ) @xmath81 ; ( e ) at ( 0,2 ) @xmath108 ; ( b ) edge [ ] node [ below ] @xmath68 ( d ) ( e ) edge [ ] node [ below ] @xmath1011 ( c ) ; ( c ) edge [ ] node [ below ] @xmath1012 ( d ) ; ( e ) edge [ ] node [ below ] @xmath698 ( b ) edge [ ] node [ above ] @xmath62 ( d ) ; let @xmath2 be a complete concrete category such that its forgetful functor preserves epis , monos and products , @xmath1 a monad on @xmath2 , @xmath74 a full subcategory of @xmath74 and @xmath61 a factorization system on @xmath2 . assume ( b@xmath3811 ) , ( b@xmath3812 ) and ( b@xmath3814 ) .
let @xmath384 be a pseudoequational @xmath1theory on @xmath74 and let @xmath95 be a pseudovariety of @xmath1algebras . then : a. let @xmath75 , we have to prove that @xmath1018 .
+ ( @xmath749 ) : let @xmath1019 with codomain @xmath966 . as @xmath966
, there exists @xmath503 such that @xmath68 factors through @xmath502 as @xmath1020 . by ( b@xmath3811 ) and ( b@xmath3814 ) we have that @xmath69 is a @xmath1algebra morphism . as @xmath1021 , then @xmath1022 , and , as @xmath384 is a pseudoequational @xmath1theory , then @xmath1023
. + ( @xmath748 ) : let @xmath389 with codomain @xmath172 . by lemma [ lemmafr ] , @xmath966 , i.e. , @xmath1019 . b. let @xmath172 be an object in @xmath57 .
+ ( @xmath748 ) : assume that @xmath97 .
we have to show that @xmath1024 .
in fact , let @xmath75 and @xmath178 .
let @xmath479 be the factorization of @xmath62 with @xmath996 and @xmath1025 .
then @xmath998 , which implies that @xmath1026 , i.e. , @xmath1024 .
+ ( @xmath749 ) : assume that @xmath1024 .
by ( r3 ) there exists an object @xmath1027 and @xmath1028 . as @xmath1024 , @xmath68 factors through some @xmath1029 as @xmath1030 .
let @xmath998 be the codomain of @xmath502 . as @xmath1021 , then @xmath1022 and @xmath1031 which implies that @xmath97 since @xmath998 .
a. define the operator @xmath436 on @xmath74 as @xmath1032 .
then @xmath436 satisfies properties i ) , ii ) and iii ) .
the proof is similar to [ exuaf ] a ) .
note that the directed union of finite objects in @xmath51 that are subobjects of @xmath339 is a distributive sublattice of @xmath339 .
b. define the operator @xmath438 on @xmath74 such that @xmath439 is the collection of all @xmath221algebra morphisms @xmath64 with domain @xmath256 and finite codomain such that @xmath1033 .
we claim that @xmath438 is a pseudoequational @xmath221theory .
non emptiness and properties ii ) and iii ) from definition [ defpet ] are proved in a similar way as in [ exuaf ] b ) .
now , to prove property i ) in definition [ defpet ] , consider a family @xmath471 in @xmath439 with @xmath100 finite such that @xmath1034 , we need to find a morphism @xmath473 such that every @xmath474 factors through @xmath68 .
in fact , let @xmath172 be the product of @xmath475 with projections @xmath476 , then , by the universal property of @xmath172 there exists a @xmath221algebra morphism @xmath477 such that @xmath478 , for every @xmath102 .
let @xmath479 be the factorization of @xmath62 in @xmath480 .
we claim that @xmath481 is a morphism in @xmath439 such that every @xmath474 factors through @xmath68 .
clearly , from the construction above , each @xmath474 factors through @xmath481 .
now , let s prove that @xmath1033 .
in fact , let @xmath482 be the codomain of @xmath481 and let @xmath1035 .
we have to prove that @xmath484 which follows from the following identity : @xmath485 where @xmath1036 is defined as @xmath1037 iff @xmath1038 .
in fact , for any @xmath850 we have that @xmath1039 implies @xmath1040 for every @xmath102 , on the other hand , if there is @xmath1041 such that @xmath1042 for every @xmath102 then @xmath1043 , i.e. , @xmath1044 for every @xmath102 ( since @xmath1045 ) , which implies that @xmath1046 ( since the order in @xmath172 is componentwise ) and the later implies that @xmath1047 ( since @xmath1048 is an embedding ) .
therefore , @xmath1039 since @xmath1041 ( i.e. , @xmath1049 ) .
+ now , for every @xmath488 and @xmath102 the composition @xmath490 belongs to @xmath210 since @xmath1050 .
as @xmath11 and @xmath100 are finite then @xmath484 because @xmath210 is a distributive lattice .
c. we have that @xmath441 .
in fact , for every @xmath75 the inclusion @xmath492 is obvious .
now , to prove that @xmath493 , let @xmath1051 with finite codomain such that @xmath495 , i.e. , @xmath1052 .
then the previous inclusion means that for every @xmath1053 there exists @xmath498 and @xmath176 such that @xmath499 .
as @xmath1054 is finite , then there exists @xmath389 such that each @xmath501 factors through @xmath68
. we will prove that @xmath502 factors through @xmath389 which will imply that @xmath503 , since @xmath384 is a pseudoequational @xmath221theory .
it is enough to show that for all @xmath853 @xmath1055 implies @xmath1056 .
in fact , assume that @xmath1055 and define @xmath1057 as @xmath507 iff @xmath1058 .
then , as @xmath509 factors through @xmath68 we have that @xmath1059 . by applying @xmath1060 to the last inequality , and using the fact that @xmath1061 , we get @xmath1062 which implies that @xmath1056 by definition of @xmath516 .
d. similar to [ exuaf ] d ) by making the obvious changes . a. define the operator @xmath436 on @xmath74 as @xmath1063 .
we claim that @xmath436 satisfies properties i ) , ii ) and iii ) .
the proof is similar to [ exuaf ] a ) .
note that the directed union of finite objects in @xmath53 that are subobjects of @xmath347 is a @xmath7vector space which is a subspace of @xmath347 .
b. define the operator @xmath438 on @xmath74 such that @xmath439 is the collection of all @xmath1algebra morphisms @xmath64 with domain @xmath87 and finite codomain such that @xmath1064 .
we claim that @xmath438 is a pseudoequational @xmath1theory .
non emptiness and properties ii ) and iii ) from definition [ defpet ] are proved in a similar way as in [ exuaf ] b ) .
now , to prove property i ) in definition [ defpet ] , consider a family @xmath1065 in @xmath439 with @xmath100 finite such that @xmath1066 , we need to find a morphism @xmath473 such that every @xmath474 factors through @xmath68 .
in fact , let @xmath172 be the product of @xmath475 with projections @xmath476 , then , by the universal property of @xmath172 there exists a @xmath1algebra morphism @xmath1067 such that @xmath478 , for every @xmath102 .
let @xmath479 be the factorization of @xmath62 in @xmath15 .
we claim that @xmath481 is a morphism in @xmath439 such that every @xmath474 factors through @xmath68 .
clearly , from the construction above , each @xmath474 factors through @xmath481 .
now , let s prove that @xmath1064 .
in fact , let @xmath482 be the codomain of @xmath481 and let @xmath1068 .
let @xmath1069 such that @xmath1070 ( this can be done since @xmath7 is injective . in fact , define @xmath1071 as zero in @xmath1072 ) and let @xmath1073 such that @xmath1074 and @xmath1075 if @xmath1076 .
note that for every @xmath1077 we have @xmath1078 .
we have to prove that @xmath484 which follows from the following identity : @xmath1079 in fact , for any @xmath1080 we have : @xmath1081 from that we get that @xmath484 since each @xmath1082 and @xmath210 is a subspace of @xmath347 . c. we have that @xmath441 .
in fact , for every @xmath75 the inclusion @xmath492 is obvious .
now , to prove that @xmath493 , let @xmath1083 with finite codomain such that @xmath495 , i.e. , @xmath1084 .
then the previous inclusion means that for every @xmath1085 there exists @xmath498 and @xmath176 such that @xmath499 .
as @xmath1086 is finite , then there exists @xmath389 such that each @xmath501 factors through @xmath68
. we will prove that @xmath502 factors through @xmath389 which will imply that @xmath503 , since @xmath384 is a pseudoequational @xmath1theory .
it is enough to show that for all @xmath1087 @xmath1088 implies @xmath515 .
in fact , assume that @xmath1088 and suppose by contradiction that @xmath1089 , then there exist @xmath1090 such that @xmath1091 , but then @xmath1088 implies @xmath1092 , since @xmath509 factors through @xmath68 , which is a contradiction since @xmath1093 .
d. similar to [ exuaf ] d ) by making the obvious changes . a. define the operator @xmath436 on @xmath74 as @xmath1094 .
we claim that @xmath436 satisfies properties i ) , ii ) and iii ) .
the proof is similar to [ exuaf ] a ) .
note that the directed union of finite objects in @xmath55 that are subobjects of @xmath1095 is a join subsemilattice of @xmath1095 .
b. define the operator @xmath438 on @xmath74 such that @xmath439 is the collection of all @xmath1algebra morphisms @xmath64 with domain @xmath87 and finite codomain such that @xmath1096 .
we claim that @xmath438 is a pseudoequational @xmath1theory .
non emptiness and properties ii ) and iii ) from definition [ defpet ] are proved in a similar way as in [ exuaf ] b ) .
now , to prove property i ) in definition [ defpet ] , consider a family @xmath1065 in @xmath439 with @xmath100 finite such that @xmath1097 , we need to find a morphism @xmath473 such that every @xmath474 factors through @xmath68 .
in fact , let @xmath172 be the product of @xmath475 with projections @xmath476 , then , by the universal property of @xmath172 there exists a @xmath1algebra morphism @xmath1067 such that @xmath478 , for every @xmath102 .
let @xmath479 be the factorization of @xmath62 in @xmath15 .
we claim that @xmath481 is a morphism in @xmath439 such that every @xmath474 factors through @xmath68 .
clearly , from the construction above , each @xmath474 factors through @xmath481 .
now , let s prove that @xmath1098 .
in fact , let @xmath482 be the codomain of @xmath481 and let @xmath1099 .
let @xmath1100 such that @xmath1070 ( this can be done since @xmath937 is an injective semilattice , see ( * ? ? ?
* lemma 1 ) ) and let @xmath1101 such that @xmath1074 and @xmath1075 if @xmath1076 .
note that for every @xmath1077 we have @xmath1102 .
we have to prove that @xmath484 which follows from the following identity : @xmath1103 in fact , for any @xmath1104 we have : @xmath1105 from that we get that @xmath484 since each @xmath1082 and @xmath210 is a join subsemilattice of @xmath1106 . c. we have that @xmath441 .
in fact , for every @xmath75 the inclusion @xmath492 is obvious .
now , to prove that @xmath493 , let @xmath1083 with finite codomain such that @xmath495 , i.e. , @xmath1107 .
then the previous inclusion means that for every @xmath1108 there exists @xmath498 and @xmath176 such that @xmath499 .
as @xmath1109 is finite , then there exists @xmath389 such that each @xmath501 factors through @xmath68
. we will prove that @xmath502 factors through @xmath389 which will imply that @xmath503 , since @xmath384 is a pseudoequational @xmath1theory .
it is enough to show that for all @xmath1110 @xmath1088 implies @xmath515 .
in fact , assume that @xmath1088 and suppose by contradiction that @xmath1089 , then there exist @xmath1111 such that @xmath1091 , but then @xmath1088 implies @xmath1092 , since @xmath509 factors through @xmath68 , which is a contradiction since @xmath1093 .
d. similar to [ exuaf ] d ) by making the obvious changes .
[ llemmafree ] let @xmath2 be a category , @xmath61 a factorization system on @xmath2 , @xmath14 a monad on @xmath2 and @xmath1112 .
assume ( b2 ) .
let @xmath567 be a local equational @xmath1theory on @xmath13
. then @xmath1113 .
[ linjth ] let @xmath2 be a category , @xmath61 a factorization system on @xmath2 , @xmath14 a monad on @xmath2 and @xmath1112 .
assume ( b1 ) and ( b2 ) .
let @xmath1115 be a local equational @xmath1theory on @xmath13 , @xmath670 .
if @xmath674 then @xmath1116 .
[ lhspclosureb ] let @xmath2 be a complete category , @xmath14 a monad on @xmath2 , @xmath61 a factorization system on @xmath2 and @xmath378 .
assume ( b1 ) , ( b2 ) and ( b3 ) .
let @xmath112 be a local equational @xmath1theory on @xmath13 .
then @xmath572 is a local variety of @xmath13generated @xmath1algebras .
a. @xmath572 is closed under @xmath59quotients : similar proof to that of proposition [ hspclosureb ] .
note that an @xmath59quotient of an @xmath13generated @xmath1algebra is @xmath13generated .
b. @xmath572 is closed under @xmath13generated @xmath60subalgebras : similar proof to that of proposition [ hspclosureb ] . c. @xmath572 is closed under subdirect products : let @xmath1117 , @xmath102 , and let @xmath482 be the subdirect product of the family @xmath557 , where @xmath1118 , @xmath102 .
let @xmath1119 then we have the following commutative diagram : + \(a ) at ( 0,0 ) @xmath82 ; ( b ) at ( 4,0 ) @xmath11 ; ( c ) at ( 2,1 ) @xmath1120 ; ( d ) at ( 0,-1.5 ) @xmath110 ; ( e ) at ( 2,-1 ) @xmath786 ; ( f ) at ( 4,-1.5 ) @xmath559 ; ( g ) at ( 4,-3 ) @xmath560 ; ( h ) at ( 6.5,-1.5 ) @xmath82 ; ( a ) edge [ ] node [ left ] @xmath112 ( d ) edge [ ] node [ above ] @xmath501 ( c ) ( d ) edge [ ] node [ above ] @xmath1121 ( e ) ( h ) edge [ ] node [ above ] @xmath1122 ( b ) edge [ ] node [ below ] @xmath561 ( g ) ; ( a ) edge [ ] node [ above ] @xmath62 ( b ) ( h ) edge [ ] node [ above ] @xmath68 ( f ) ( d ) edge [ ] node [ below ] @xmath69 ( f ) edge [ ] node [ below ] @xmath1123 ( g ) ( f ) edge [ ] node [ right ] @xmath563 ( g ) ; ( e ) edge [ ] node [ above ] @xmath1124 ( f ) ( c ) edge [ ] node [ above ] @xmath1048 ( b ) ( b ) edge [ ] node [ right ] @xmath564 ( f ) ; + where : 1 . the two right triangles are obtained from the construction of @xmath482 , 2 .
@xmath479 is the factorization of @xmath62 , 3 .
@xmath1123 is obtained from the property that @xmath1125 , i.e. , @xmath1126 , @xmath565 , 4 .
@xmath69 is obtained from the morphisms @xmath1123 by using the universal property of the product @xmath558 , and 5 .
@xmath974 is the factorization of @xmath69 .
+ then , by the universal property of the product , we have that @xmath1127 , where @xmath1128 and @xmath1129 .
hence , by uniquenes of factorization we have that there is an isomorphism @xmath978 such that @xmath1130 , which implies that @xmath1131 , i.e. , @xmath1132 .
[ lpropvartoeq ] let @xmath2 be a complete category , @xmath61 a factorization system on @xmath2 , @xmath14 a monad on @xmath2 and @xmath378 .
assume ( b1 ) , ( b3 ) and ( b4 ) .
let @xmath103 be a local variety of @xmath13generated @xmath1algebras .
then @xmath1133 for some local equational @xmath1theory @xmath112 on @xmath13 .
a. let @xmath1134 be the collection of all @xmath1algebra morphisms , up to isomorphism , in @xmath59 with domain @xmath82 and codomain in the variety @xmath103 . by ( b4 )
, @xmath1135 is a set .
put @xmath1136 and let @xmath1137 be the @xmath532th projection .
then we have the following commutative diagram in @xmath15 : + \(a ) at ( 2,0 ) @xmath110 ; ( b ) at ( 6,0.5 ) @xmath1138 ; ( c ) at ( 0,0.5 ) @xmath82 ; ( d ) at ( 4,0.5 ) @xmath786 ; ( a ) edge [ ] node [ below ] @xmath1139 ( d ) ; ( c ) edge [ ] node [ below ] @xmath112 ( a ) ; ( c ) edge [ ] node [ above ] @xmath698 ( d ) ; ( d ) edge [ ] node [ above ] @xmath712 ( b ) ; ( c ) edge [ bend left , looseness=0.7 ] node [ above ] @xmath474 ( b ) ; + where @xmath1140 is obtained from the universal property of the product @xmath829 and @xmath1141 is the factorization of @xmath698 , i.e. , @xmath1142 and @xmath1143 . observe that @xmath1144 since it is a subdirect product of elements in @xmath103 .
+ _ claim : _ @xmath567 is a local equational @xmath1theory on @xmath13 . + let @xmath568 .
we have to prove that there exists @xmath107 such that @xmath1145 .
in fact , we have the following commutative diagram : + \(b ) at ( 2,0 ) @xmath786 ; ( c ) at ( 0,1.5 ) @xmath82 ; ( d ) at ( 0,0 ) @xmath110 ; ( e ) at ( 6.5,1.5 ) @xmath82 ; ( f ) at ( 6.5,0 ) @xmath110 ; ( g ) at ( 4,0 ) @xmath1146 ; ( g ) edge [ ] node [ below ] @xmath1147 ( f ) ( d ) edge [ ] node [ below ] @xmath200 ( b ) ; ( c ) edge [ ] node [ left ] @xmath112 ( d ) ( e ) edge [ ] node [ right ] @xmath112 ( f ) ( c ) edge [ ] node [ right ] @xmath1148 ( g ) ; ( c ) edge [ ] node [ above ] @xmath69 ( e ) ( b ) edge [ ] node [ below ] @xmath563 ( g ) ; + where @xmath1149 is the factorization of @xmath1150 and @xmath482 is the codomain of @xmath1151 . from that we have then that @xmath482 is an @xmath13generated @xmath60subalgebra of @xmath1144 .
hence @xmath1152 and therefore @xmath1153 and @xmath1154 for some @xmath565 .
finally , commutativity of the triangle follows from the definition of @xmath875 above .
therefore , @xmath112 is a local equational @xmath1theory on @xmath13 .
b. let us prove that @xmath1133 .
+ @xmath1155 : let @xmath84 such that @xmath1156 .
since @xmath172 is @xmath13generated , there exists @xmath1157 . as @xmath1156 , the morphism @xmath1158 factors through @xmath112 as @xmath1159 . as we have that @xmath1160 then , by ( b1 ) and lemma [ lemmafs ] , we have that @xmath1161 , and hence @xmath1162 since it is an @xmath59quotient of @xmath1144 .
+ @xmath1163 : let @xmath84 such that @xmath1162 .
let @xmath178 , then we have the following commutative diagram : + \(a ) at ( 0,0 ) @xmath82 ; ( b ) at ( 3.5,0 ) @xmath110 ; ( c ) at ( 3.5,1.5 ) @xmath786 ; ( d ) at ( 3.5,3 ) @xmath1164 ; ( e ) at ( 0,3 ) @xmath81 ; ( a ) edge [ ] node [ below ] @xmath112 ( b ) edge [ ] node [ left ] @xmath1165 ( d ) ; ( a ) edge [ ] node [ left ] @xmath62 ( e ) ( c ) edge [ ] node [ right ] @xmath712 ( d ) ; ( b ) edge [ ] node [ right ] @xmath200 ( c ) ( d ) edge [ ] node [ above ] @xmath1166 ( e ) ; + where @xmath1167 is the factorization of @xmath62 with @xmath1168 and @xmath1169 , which implies that @xmath1170 since it is an @xmath13generated @xmath60subalgebra of @xmath1162 . therefore , @xmath1171 and @xmath1165 for some @xmath102 .
hence the factorization of @xmath62 through @xmath112 follows from the definition of @xmath112 ( see i ) above ) which implies that @xmath1156 .
[ [ proof - of - local - birkhoffs - theorem - for - finite - mathsftalgebras ] ] proof of local birkhoff s theorem for finite @xmath1algebras ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ [ lhspclosurer ] let @xmath2 be a concrete complete category such that its forgetful functor preserves epis , monos and products , @xmath1 a monad on @xmath2 , @xmath378 and @xmath61 a factorization system on @xmath2
. assume ( b@xmath3811 ) to ( b@xmath3813 ) .
let @xmath584 be a local pseudoequational @xmath1theory on @xmath13 .
then @xmath1172 is a local pseudovariety of @xmath13generated @xmath1algebras .
a. @xmath1172 is closed under @xmath59quotients : similar proof to that of proposition [ hspclosurer ] .
note that an @xmath59quotiend of an @xmath13generated algebra is also @xmath13generated .
b. @xmath1172 is closed under @xmath13generated @xmath60subcoalgebras : similar proof to that of proposition [ hspclosurer ] . c. @xmath1172 is closed under finite subdirect products : let @xmath100 be a finite set and let @xmath1173 , @xmath102 .
let @xmath482 be the subdirect product of the family @xmath557 , where @xmath1118 , @xmath102 .
let @xmath1119 then we have the following commutative diagram : + \(a ) at ( 0,0 ) @xmath82 ; ( b ) at ( 4,0 ) @xmath11 ; ( c ) at ( 2,1 ) @xmath1120 ; ( d ) at ( -2,-3 ) @xmath1174 ; ( e ) at ( 2,-1 ) @xmath786 ; ( f ) at ( 4,-1.5 ) @xmath559 ; ( g ) at ( 4,-3 ) @xmath560 ; ( h ) at ( 0,-1.5 ) @xmath1010 ; ( i ) at ( 6.5,-1.5 ) @xmath82 ; ( a ) edge [ ] node [ left ] @xmath561 ( d ) edge [ ] node [ right ] @xmath83 ( h ) edge [ ] node [ above ] @xmath501 ( c ) ( h ) edge [ ] node [ above ] @xmath1121 ( e ) ( i ) edge [ ] node [ above ] @xmath1122 ( b ) edge [ ] node [ below ] @xmath561 ( g ) ; ( a ) edge [ ] node [ above ] @xmath62 ( b ) ( h ) edge [ ] node [ below ] @xmath69 ( f ) ( d ) edge [ ] node [ below ] @xmath1123 ( g ) ( f ) edge [ ] node [ right ] @xmath563 ( g ) ( h ) edge [ ] node [ right ] @xmath1175 ( d ) ( i ) edge [ ] node [ above ] @xmath68 ( f ) ; ( e ) edge [ ] node [ above ] @xmath1124 ( f ) ( b ) edge [ ] node [ right ] @xmath564 ( f ) ( c ) edge [ ] node [ above ] @xmath1048 ( b ) ; + where : 1 .
the two right triangles are obtained from the construction of @xmath482 , 2 .
@xmath479 is the factorization of @xmath62 , 3 .
@xmath1123 is obtained from the property that @xmath1176 , i.e. , @xmath1177 , @xmath565 and @xmath1178 , 4 .
@xmath1179 is obtained from the morphisms @xmath1178 by using the property that @xmath1180 is a local pseudoequational @xmath1theory on @xmath13 , i.e. , @xmath1181 , @xmath565 , 5 .
@xmath69 is obtained from the morphisms @xmath1182 by using the universal property of the product @xmath558 , and 6 .
@xmath974 is the factorization of @xmath69 .
+ then , by the universal property of the product , we have that @xmath1183 .
now , as @xmath1184 and @xmath1129 , by uniquenes of factorization we have that there is an isomorphism @xmath978 such that @xmath1185 , which implies @xmath1186 , i.e. , @xmath1187 .
let @xmath2 be a concrete complete category such that its forgetful functor preserves epis , monos and products , @xmath1 a monad on @xmath2 , @xmath378 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) and ( b@xmath3813 ) .
let @xmath95 be a local pseudovariety of @xmath13generated @xmath1algebras .
then @xmath1188 is a local pseudoequational @xmath1theory on @xmath13 .
a. let @xmath100 be a finite set and @xmath1190 , @xmath102 .
let @xmath101 be the codomain of @xmath388 .
let @xmath710 with projections @xmath994 .
now , by definition of @xmath172 , there exists @xmath178 such that @xmath995 .
let @xmath479 be the factorization of @xmath62 with @xmath996 and @xmath997 .
we have that @xmath998 since it is the subdirect product of @xmath1191 . hence , @xmath1192 and every @xmath388 factors through it .
b. follows from the property that @xmath95 is closed under @xmath59quotients . c.
let @xmath1193 with codomain @xmath97 , and @xmath589 .
let @xmath395 be the factorization of @xmath396 such that @xmath1194 and @xmath1006 .
then @xmath998 since it is an @xmath13generated @xmath60subcoalgebra of @xmath97 , which implies @xmath1195 .
let @xmath2 be a concrete complete category such that its forgetful functor preserves epis , monos and products , @xmath1 a monad on @xmath2 , @xmath378 and @xmath61 a factorization system on @xmath2 .
assume ( b@xmath3811 ) to ( b@xmath3813 ) .
let @xmath584 be a local pseudoequational @xmath1theory on @xmath13 and let @xmath95 be a local pseudovariety of @xmath13generated @xmath1algebras . then : a. ( @xmath749 ) : let @xmath1198 with codomain @xmath1199 .
as @xmath1200 , there exists @xmath1201 such that @xmath68 factors through @xmath502 as @xmath1020 . by ( b@xmath3812 ) and
( b@xmath3814 ) we have that @xmath69 is a @xmath1algebra morphism . as @xmath1021 , then @xmath1022 , and , as @xmath584 is a pseudoequational @xmath1theory , then @xmath1202 .
+ ( @xmath748 ) : let @xmath587 with codomain @xmath172 . using lemma [ lemmafr ] with @xmath1114 , @xmath1203 , i.e. , @xmath1198 .
b. let @xmath172 be an @xmath13generated algebra in @xmath57 .
+ ( @xmath748 ) : assume that @xmath97 .
we have to show that @xmath1204 .
in fact , let @xmath178 and @xmath479 be the factorization of @xmath62 with @xmath996 and @xmath1025 .
then @xmath998 since it is an @xmath13generated @xmath60subcoalgebra of @xmath97 , which implies that @xmath1192 , i.e. , @xmath1204 .
+ ( @xmath749 ) : assume that @xmath1204 .
since @xmath172 is an @xmath13generated @xmath1algebra , there exists @xmath1205 . as @xmath1204 , @xmath68 factors through some @xmath1206 as @xmath1030 .
let @xmath998 be the codomain of @xmath502 . as @xmath1021 , then @xmath1022 and @xmath1031 which implies that @xmath97 since it is an @xmath59quotient of @xmath998 . | the purpose of the present paper is to show that : @xmath0 we consider algebras for a monad @xmath1 on a category @xmath2 and we study ( pseudo)varieties of @xmath1algebras .
pseudovarieties of algebras are also known in the literature as varieties of finite algebras .
two well known theorems that characterize varieties and pseudovarieties of algebras play an important role here : birkhoff s theorem and birkhoff s theorem for finite algebras , the latter also known as reiterman s theorem .
we prove , under mild assumptions , a categorical version of birkhoff s theorem for ( finite ) algebras to establish a one to
one correspondence between ( pseudo)varieties of @xmath1algebras and ( pseudo)equational @xmath1theories .
now , if @xmath3 is a category that is dual to @xmath2 and @xmath4 is the comonad on @xmath3 that is the dual of @xmath1 , we get a one to one correspondence between ( pseudo)equational @xmath1theories and their dual , ( pseudo)coequational @xmath4theories .
particular instances of ( pseudo)coequational @xmath4-theories have been already studied in language theory under the name of `` varieties of languages '' to establish eilenberg type correspondences .
all in all , we get a one to one correspondence between ( pseudo)varieties of @xmath1algebras and ( pseudo)coequational @xmath4theories , which will be shown to be exactly the nature of eilenberg type correspondences . |
traces of heavy metals in helium - dominated atmospheres ( spectral types dbz , dz ) have been known since almost one hundred years in fact , van maanen 2 , one of the three `` classical '' white dwarfs belongs to this class .
although the majority of wds has hydrogen - dominated atmospheres ( da ) , the story of metal pollution in these objects is comparatively recent . consisting for many years only of the prototype g74 - 7 @xcite ,
the second and third member of the class were only discovered in 1997 @xcite . since
then the number has grown very fast to more than 100 today , including the ones reported here .
this difference between helium and hydrogen atmospheres is purely observational bias : at the same abundance the equivalent width of the caii resonance lines in helium is a factor of hundred or more larger than in a da , due to the much larger transparency of helium at temperatures below 20000 k. large telescopes like the vlt and keck were therefore essential for this development .
metals in these objects can not be primordial .
gravitational settling with diffusion timescales of days to millions of years are always shorter than the evolutionary timescales , and an external source is required to replenish the heavy elements .
historically the standard explanation has been accretion from interstellar matter , although severe problems notably the large underabundance of hydrogen in the accreted matter and the absence of dense clouds in the solar neighborhood were known and discussed since decades . with the detection of an infrared excess around a significant fraction of the polluted objects by the spitzer space telescope and other ground - based ir telescopes the explanation has now completely shifted towards the accretion from circumstellar dust in a debris disk , originating from the tidal disruption of rocky material from a planetary system .
this offers a completely new method to study the elemental composition of extrasolar planetary systems by relating the photospheric abundances in the white dwarf to the abundances in the accreted matter using the equations of the accretion / diffusion scenario ( e.g. * ? ? ?
* ) . for helium - rich and cool hydrogen - rich white dwarfs with extended outer convection zones
the diffusion timescales can be of the order of a million years and the current status remains uncertain beginning accretion , equilibrium between accretion and diffusion , or accretion already ended with corresponding uncertainties for the composition of the accreted matter . we ( pi boris gnsicke ) have therefore initiated a program to observe a large number of da white dwarfs with effective temperatures between 17000 and 25000 k with the cos spectrograph on the hst ( implemented as a snapshot program ) .
these das have no outer convection zones , the diffusion timescales are extremely short and equilibrium between accretion and diffusion is guaranteed .
the atmospheric abundances together with diffusion velocities calculated from the atmosphere models directly give the diffusion flux for each observed heavy element and thus the relative abundances in the accreted matter .
the wavelength range observed was 1130 - 1435 , with typical exposure times of 1000 s. this range includes strong lines of si and c , and fairly strong lines of other important elements like o , al , s , fe .
results for four heavily polluted objects have been published in @xcite and details of the observations , parameter determinations , and synthetic spectra calculations can be found there .
here we report on 78 additional das , but include the first four in the statistics and figures .
all objects show interstellar absorption lines of si and c , and most additionally from n , s , o , fe .
the most direct indication is the siii1260 line and the absence or weakness of siii1265 .
the latter line has the larger gf value , but its lower level is 0.036 ev above the real ground level ; it is always the stronger line in the stellar photosphere , but never appears as interstellar line in our sample . in most cases the photospheric lines ( if present )
are also shifted by 0.1 - 0.3 relative to the ism lines .
the spectral resolution of @xmath1 km / s does not allow to identify individual components in the ism lines .
however , the equivalent width of the ism siii1260 line ranges from 21 to 184 m , corresponding to column densities of @xmath2 @xmath3 , if interpreted as single line on the linear part of growth ( e.g. * ? ? ?
these column densities are typical values for the local interstellar cloud ( lic ) surrounding our solar system @xcite ; very likely this is the only absorbing ism material between us and most of our sample .
49 of 82 or 60% of our objects show photospheric si , 18 additionally c , and 8 further metals .
taken at face value this number is surprisingly high , as previous estimates ( e.g. * ? ? ?
* ) put it much lower at 20 - 25% , and the clear signature of the connection with a planetary system , the infrared excess of a debris disk , is only found in about 1% of white dwarfs @xcite . since the diffusion timescales are only a few days , the natural assumption would be that all these objects are currently accreting , with accretion fluxes for silicon within a factor of 10 of @xmath4 g / s .
however , before accepting such a conclusion , there is a caveat . @xcite
have shown that some elements , in particular si and c , can be supported by radiative acceleration in the atmosphere even at @xmath0 as low as 20000 k. this would mean that no current accretion is needed , or that the accretion flows can be much smaller than calculated from the abundances . fig .
[ figsi ] shows the si abundances as a function of a parameter calculated from the photon flux ( @xmath5 ) minus the surface gravity @xmath6 .
we take this as a qualitative measure of the radiative acceleration on si , acting against the downward gravitation . unfortunately ,
@xcite publish details only for one model with @xmath7 k , @xmath8 , but predicted abundances for four real stars can be extracted from a companion paper @xcite .
all upper limits fall below these prediction .
the interpretation in this scenario is that these stars have never had si in their outer layer , or lost it long ago either by downward gravitational settling or expulsion via a stellar wind .
the other side of the coin is the conclusion that all objects currently showing photospheric si must have accreted fairly recently ( after levitation became effective enough to prevent settling ) or are accreting right now .
there are a large number of abundance measurements below the equilibrium line , which can not be understood taking the theoretical and observational data at face value .
we note , however , that a better agreement within the abundance errors for these objects could be achieved by shifting the prediction downward by about 0.4 dex .
a reason for such an apparent discrepancy can easily be found .
3 in @xcite predicts very strong abundance gradients within the atmosphere , whereas our analysis only uses homogeneous distributions without radiative levitation . a real test of the levitation predictions could only be done by incorporating the effect in the model atmospheres and calculate the expected spectra .
whatever the explanation of such a shift , it would explain the majority of the si observations with si accretion rates @xmath9 100 g / s .
15 - 20 objects would remain above the predictions .
these could only be explained by ongoing accretion from circumstellar material and the accretion rates as determined by the simple description without levitation would give correct results .
there are much fewer objects with c abundance determinations .
the data in @xcite strongly suggest that carbon should not be supported in the atmosphere at lower @xmath0 .
yet there are 18 cases of observed photospheric c , similar to the number of clear accretion cases in the discussion of si .
again , the only viable explanation for these seems to be ongoing accretion .
[ ratiocsi ] shows the carbon to silicon ratio or upper limits , as well as those numbers for the sun and bulk earth .
there are some values close to the solar / ism value ( possibly accretion from a low mass companion ? ) but the majority are a factor of 10 or more below solar .
@xcite found that the c / si ratio of all four objects was close to the bulk earth value ( in fact the four lowest points in the fig .
[ ratiocsi ] are from that paper ) .
it should be noted here that the si abundance is generally much lower in the new objects and with a bulk earth ratio carbon would not be seen in these objects ( as is shown by the many upper limits ) .
the empty lower left part of the figure is thus probably observational bias .
that the upper right part is also empty indicates that very high accretion rates occur only with carbon depleted matter .
the accretion rates in the 15 - 20 objects clearly above the predicted levitation abundances are also too high to be explained by accretion from the thin interstellar matter in the local bubble around the sun . in view of the established connection between infrared excess , circumstellar dust , and metal pollution in the stellar atmosphere
, the most likely explanation seems to be accretion from a debris disk
. however , that is not necessarily true for the majority of polluted daz .
if the levitation calculations are qualitatively correct , the amount of si necessary to explain the observations can be extremely small .
the total mass in a 20000 k , @xmath8 atmosphere above @xmath10 is approximately @xmath11 g . using data from fig .
3 in @xcite the maximum abundance in the outer layers of such a model is n(si)/n(h ) = @xmath12 , and therefore the mass of si is smaller than @xmath13 g. the cross section for eddington accretion from interstellar matter , which is the minimum we would expect , is @xmath14 with white dwarf mass and radius @xmath15 and @xmath16 , gravitational constant @xmath17 and space velocity @xmath18 ( assumed to be 30 km / s ) .
assuming the white dwarf crosses just one tiny cloud like our local interstellar cloud ( lic ) with a si column density of @xmath19 @xmath3 , it would sweep up @xmath20 g of si .
that is orders of magnitude more than needed to explain the observations , which would in this scenario diffuse downward , except for the tiny fraction supported by radiative levitation .
all 82 objects show interstellar absorption , very likely due to the local interstellar cloud ( lic ) surrounding the solar system . more importantly
, 60% also show photospheric si , and a smaller fraction also c and a few other metals . if the predictions of radiative levitation are correct , the majority could be explained by very low accretion rates that could be possible even within our local bubble . that leaves about 20% definitely accreting from circumstellar material , in agreement with previous estimates .
the carbon to silicon ratio is very low ( similar to b.e .
values ) in the cases of high accretion rates , and shows a range of values for the lower rates , although mostly below the solar ratio . | a total of 82 da white dwarfs have been observed with the cosmic origins spectrograph on the hubble space telescope in a snapshot program .
the targets were selected to be in the @xmath0 range from 17000 - 25000 k , where optical metal lines become weak and difficult to detect . because of the strong si , c , and o resonance lines in the uv , this survey has a sensitivity that is comparable to that of the keck / vlt searches for caii k in cooler white dwarfs .
these objects also have no convection zone and thus very short diffusion timescales , assuring that accretion is currently ongoing .
the spectra have high resolution and in most cases fairly good s / n . about 60% of them show photospheric metal pollution , predominantly of si , but in some cases additional metals are present .
we report the results of a preliminary analysis and discuss the sources of the accreted matter and the possible rle of radiative levitation . |
hierarchical structure formation predicts that regions with the highest density are the sites of the earliest galaxy formation .
the low - redshift descendants of such regions ( virialized galaxy clusters ) have been extensively studied and various relationships between galaxy properties and environmental density have been found ( e.g. , morphology / star - formation - density relation ; * ? ? ?
* ; * ? ? ?
. however , the dominant physical processes that drive galaxy evolution in different environments and cause the observed present - day relationships remain poorly understood .
several studies have demonstrated that the red sequence that characterizes present - day clusters is in place as early as @xmath13 and that star - formation activity in these dense environments must have peaked at earlier epochs @xcite .
in addition , studies of a handful of clusters at redshifts @xmath14 reveal that star - formation is enhanced in dense regions : the slope of the star - formation
density relation observed at low - redshift appears to decrease and perhaps `` reverse '' at @xmath15 , beyond which redshift cluster progenitors are expected to show enhanced star - formation relative to the field ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the evolution at even earlier epochs remains uncertain , primarily because of the lack of good protocluster candidates .
identifying protocluster regions at early epochs is important both to our understanding of large scale structure evolution and to the study of galaxy evolution in dense environments .
however , finding un - virialized high - redshift ( @xmath16 ) `` protoclusters '' is very challenging . at high redshift , identifying robust overdensities through purely photometric means is difficult because the commonly - used broad - band selection techniques result in broad redshift distributions ( and many interloper populations ) that significantly reduce the observed density contrast .
spectroscopy of large numbers of faint continuum - selected candidates requires expensive allocations of large telescope time . unlike low - redshift ( @xmath17 ) clusters
, high - redshift protoclusters do not show detectable x - ray emission ; at @xmath16 red sequences may not exist , or if they do , are difficult to identify observationally . despite these challenges ,
various heroic efforts have resulted in a heterogeneous sample of robust high - redshift protoclusters .
the majority at @xmath16 have been found by targeting rare sources ( e.g. , radio galaxies or ly@xmath8 blobs * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) ; a handful of structures have been found in `` blank - field '' surveys @xcite .
recently , @xcite identified a number of protocluster candidates at @xmath18 in the cosmos field using photometric redshifts , but only two have been thus far confirmed using spectroscopic redshifts .
the overall number of confirmed protoclusters ( i.e. , @xmath19 at @xmath20 ) is still too low to allow for robust comparisons to the predictions of hierarchical structure formation models or for environmental studies of galaxy properties at different redshifts @xcite . in this work
we report the discovery of an overdense region ( roughly 0.3 deg@xmath12 in size ) containing multiple protocluster candidates lying within the botes field of the noao deep wide - field survey ( ndwfs : * ? ? ?
the presence of a protocluster was first signaled by the distribution of uv - continuum - selected galaxies . here ,
we present the identification of robust ly@xmath8 line - emitting galaxy candidates in the field and use these to map and weigh the large - scale structure . in 2 , we describe the initial search technique that resulted in identifying a protocluster candidate and thus selecting the field for further study . in 3 , we describe new data that we obtained to map the distribution of ly@xmath8 emitters in the field ; we detail the selection of the candidates and assess their robustness . in
4 we describe the distribution of galaxies in the structure and assess the overdensities of the protocluster candidates in the field .
we conclude in 5 . throughout
, we use the wmap7 cosmology @xmath21 @xcite ; at @xmath0 , the angular scale is 7.347kpc/(physical ) .
distance scales are presented in units of comoving mpc unless noted otherwise .
all magnitudes are given in the ab system @xcite .
we have been undertaking a search for uv - luminous star - forming galaxies at @xmath22 in the botes field of the ndwfs .
we refer the reader to @xcite and @xcite for details regarding the search , but briefly describe the procedure here .
we applied a standard lyman - break galaxy color - selection technique to the ndwfs @xmath23 data and selected candidates based on the following criteria ( same as lee et al .
2011 , but in ab mag ) : @xmath24 for sources undetected in the @xmath25-band , we used the @xmath26 limit .
a subset of the candidates was observed spectroscopically using the deimos instrument at the keck ii telescope of the w. m. keck observatory ; these observations confirmed the redshift distribution of the sample and characterized the selection function @xcite .
these spectroscopic observations also resulted in the discovery of five galaxies lying at redshifts @xmath27 , within 1 mpc ( projected physical distance ) from one another ( see figures 11 & 12 and 5.1 in * ? ? ? * ) .
we interpreted this alignment of galaxies as signaling the presence of a coherent large scale structure and set out to measure its extent and spatial overdensity using ly@xmath8 emitting galaxies as a tracer population .
in may and june 2012 , we obtained deep imaging of the candidate protocluster field using the mosaic 1.1 camera on the mayall 4-m telescope of the kitt peak national observatory .
images were obtained through both broad - band ( @xmath23 ) and narrow - band ( wrc4 ) filters centered on the field ndwfsj1426 + 3236 .
the wrc4 filter ( designed to sample emission in wolf - rayet stars ; kpno filter # k1024 ) has a central wavelength ( in the kpno 4 m f/3.1 prime - focus corrector beam ) of 5819 and a full - width at half maximum of 42 ; it is therefore well placed to sample ly@xmath8 emission at @xmath28 ( i.e. , 25 comoving mpc ) .
we used individual exposure times of [ 20,20,10,10 ] minutes for the [ @xmath25,wrc4,@xmath29,@xmath30 bands respectively and dithered the telescope between exposures by @xmath31 arcmin . after discarding frames taken with delivered image quality @xmath32 , the effective total exposure times of the new imaging are 8.3 , 3.0 , 5.2 , and 0.8 hours for the wrc4 , @xmath25 , @xmath29 , and @xmath33 bands respectively . for each frame , we updated the astrometry ( employing the iraf task msccmatch ) using stars identified in the sloan digital sky survey dr7 catalog , and reprojected it to a common tangent point with a pixel scale of 0.258/pixel .
the same procedure was repeated for the ndwfs frames after recomputing the mosaic camera image distortions using a large catalog of sdss dr7 sources .
the @xmath25 , @xmath29 , and @xmath33 band images are photometrically tied to the existing ndwfs images .
the relative intensity scale of each reprojected frame was determined with respect to the ndwfs data using the mscimatch task in the iraf mscred package .
all frames of a given band were combined to construct the final image stack using a weighted average , with the relative weight inversely proportional to the variance of sky noise measured in the reprojected frames .
we masked regions near the field borders with less than 20% of the maximum exposure time and areas near bright saturated stars and their diffraction spikes .
the effective area of the final mosaic is 0.317 deg@xmath12 .
the photometric calibration for the @xmath34 images is referenced to the ndwfs images . for the wrc4 band
, we determined the photometric zero point using spectrophotometric standard stars observed during the run . combining the new data with the existing ndwfs data results in effective exposure times of 8.3 , 5.3 , 6.9 , and 4.1 hours for the wrc4 , @xmath25 , @xmath29 , and @xmath33 bands respectively , resulting in @xmath35 limiting magnitudes of
24.91 , 26.96 , 26.34 , and 25.55 ab mag ( measured on blank sky regions within 2@xmath36 diameter circular apertures ) .
the delivered image quality is @xmath37 in the final stacks .
we estimate the detection completeness by inserting into the stacked images artificial sources with similar sizes and morphologies as real high - redshift galaxies .
we find that the 90% , 50% , and 10% completeness limits for source detection are 23.82 , 24.22 , 24.64 mag , respectively , for the wrc4 band , and 24.07 , 24.78 , 25.26 mag for the @xmath33- band .
our primary goal is to investigate the spatial extent of the presumed galaxy overdensity at @xmath0 as traced by ly@xmath8-emitting galaxies ( laes ) .
we identify candidate laes as sources that show excess emission in the wrc4 band , i.e. , unusually blue narrow - band - to - broad - band colors . in order to understand the sample selection criteria , we synthesized the ( @xmath40 ) and ( @xmath42 ) colors of model laes with a large range in rest - frame uv continuum slope , ly@xmath8 emission line equivalent width , and ly@xmath8 luminosity .
first , we created a model lae spectrum as follows .
we used the stellar population synthesis model of @xcite to compute the continuum spectrum of a galaxy with a constant star formation history observed at the population age of 100 myr , assuming a salpeter initial mass function and solar metallicity .
we accounted for attennuation by intergalactic hydrogen using the model of @xcite .
we adopted the dust extinction model of @xcite and varied the reddening parameter @xmath43 until the predicted @xmath44 color approximately matched the observed value @xmath45 as shown in figure [ lae_selection ] ( left panel ) .
to the continuum spectrum , we added a ly@xmath8 emission line with a gaussian line profile centered at 1215.67 and an intrinsic line width of 3 .
exact values assumed for the line width and reddening parameter are not important as long as they reasonably reproduce the observed galaxy colors and line fwhm . model spectra were generated for a range of ly@xmath8 rest - frame equivalent widths , @xmath46 , and luminosities and used to predict @xmath42 colors ( see figure [ lae_selection ] , left panel ) .
the color criterion @xmath47 corresponds to @xmath48 at @xmath49 .
figure [ lae_selection ] also shows the expected @xmath50 and @xmath51 colors at different extinction / reddening values . at a fixed reddening value , large open triangles
( diamonds , circles ) show the ly@xmath8 luminosities of , from top to bottom , @xmath52 , @xmath53 , @xmath54 , and @xmath55 at the continuum @xmath29-band magnitude of 25.5 ( 25.2 , 25.8 ) mag .
the median @xmath29-band magnitude is 25.5 , with the majority at @xmath560.6 mag .
the limiting ly@xmath8 luminosity of our survey is @xmath57 .
we find 65 lae candidates within the total survey area of 0.317 deg@xmath12 , corresponding to an average surface density of @xmath58 arcmin@xmath59 .
ten of the lae candidates are only detected in the wrc4 band and have no @xmath29 or @xmath33-band continuum detections ( shown as triangles in figure [ lae_selection ] .
approximate ly@xmath8 luminosities corresponding to their wrc4-band magnitudes are indicated on the top abscissa ; the majority of our sources have @xmath60 and in the wrc4 magnitude range of @xmath61 , while there are six laes brighter than @xmath62 ( @xmath63 ) .
in addition to the lae candidates , we also used the deeper @xmath23 broad - band imaging to identify a sample of lyman - break star - forming galaxies ( lbgs ) at @xmath65 using the selection criteria described in equation ( 1 ) .
the reliability of @xmath64 lbg candidates selected according to these criteria has been investigated by @xcite .
we identified 428 lbg candidates within our field , i.e. , an observed candidate surface density of [email protected] arcmin@xmath59 .
accounting for the possible contamination by interlopers ( measured by * ? ? ?
* to be @xmath67% ) , this suggests an lbg surface density in the field of [email protected] arcmin@xmath59 , comparable to that reported in @xcite .
the broad - band colors of the lbg candidates are shown in figure [ lbg_selection ] as black dots within selection window .
their angular distribution is shown in figure [ distribution ] ( grey circles ) .
although the bulk of our sample awaits spectroscopic confirmation , four of the five spectroscopically confirmed uv - continuum - selected lyman break galaxies are members of our lae candidate list .
photometrically - selected lae samples can be contaminated by low - redshift interloping galaxies where , e.g. , emission lines of [ o iii ] and [ o ii ] fall within the bandpass of the narrow - band filter .
low - redshift interlopers at @xmath68 with strong [ oiii]@xmath695008 emission , or at @xmath70 with strong [ oii]@xmath693729 emission , will exhibit a @xmath71 color excess and can potentially contaminate the lae candidate sample .
although they may satisfy the @xmath71 criterion , they may be discriminated by the @xmath72 selection criteria .
figure [ lbg_selection ] shows the @xmath72 and @xmath51 color - color diagram of all detected sources in the field ( dots ) ; the subset of 17 lae candidates detected in both @xmath29 and @xmath33 bands are also shown ( red filled circles ) .
also shown are spectroscopically confirmed galaxies from the agn and galaxy evolution survey ( ages ; * ? ? ?
* ) that lie within the redshift ranges @xmath68 and @xmath70 ( blue triangles ) . although the ages survey targeted bright galaxies ( @xmath73 vega mag ) , none of them have colors that overlap with the lae sample .
two of the 17 sources lie between the diagonal lbg selection criterion boundary and the low - redshift galaxy sequence , but there is no overlap between the populations . accounting for upper limits , we find that only three of the 55 laes with continuum detection lie within 0.2 mag of the low - redshift galaxy sequence .
we therefore conclude that the contamination is no more than 6% .
while laes are typically powered by star - formation , some of the more luminous emitters could be agn ( e.g. , * ? ? ?
while this is not an issue for tracing the structures , we briefly consider whether there is evidence for agn within our lae candidate sample .
the x - ray data that exist in the field ( from the chandra xbotes survey ; * ? ? ?
* ) are too shallow to place firm constraints on agn contamination for the bulk of the sample .
however , none of of the lae candidates are detected in the x - ray .
all but two lae candidates are covered by existing _ spitzer _
mips 24@xmath74 m imaging . among them , only one is clearly detected .
visual inspection of the morphologies show that most of them are spatially extended ( @xmath75 or 1.2 ) and not dominated by a point source component that may suggest the presence of an agn .
two have existing rest - frame uv spectra @xcite , but neither shows c iv or n v emission indicative of agn .
we therefore conclude that the contamination of the lae candidate population by luminous agn is not significant .
some guidance may also be found from other work .
@xcite , whose survey is the closest match to ours in terms of redshift and imaging depth , did not find any clear evidence of contamination in their @xmath76 lae spectroscopic sample . after similar narrow - band searches of laes around high - redshift radio galaxies and their spectroscopic followup
, @xcite reported that contamination rate averaged over all fields is @xmath77% ( in some fields , they found zero contamination ) .
while we can not be certain without spectroscopy of the candidates , we conclude that the selection of the lae candidates is robust and that most , if not all , lie at @xmath38 . in
what follows , we use the spatial distribution of laes to estimate overdensities ; any low-@xmath78 interloper contamination , if randomly distributed , will result in an underestimate of the true overdensity .
in this section , we discuss the angular distribution of the galaxies within the field .
we consider two samples separately : ( 1 ) the 65 lae candidates supplemented by the spectroscopically confirmed galaxies ( which adds one extra galaxy to this subset ) ; and ( 2 ) the 428 lbg galaxy candidates .
the distributions of these two subsets are shown in the left panel of figure [ distribution ] .
the lae distribution within the field ( which covers a volume of @xmath2 mpc at @xmath49 ; figure [ distribution ] ) is highly inhomogeneous . in particular , there are three regions ( denoted as `` ne '' , `` nw '' , and `` s '' groups ) each containing @xmath79 laes within a 8 mpc ( or 3.8 ) radius circle . given the observed average lae surface density in this field of 0.057 arcmin@xmath59 ( @xmath80 ) , the expected number of laes within a 3.8 radius circle is @xmath81 . hence , these three regions have surface densities that are factors of @xmath82 higher than the average in this field , or surface overdensities of @xmath83 . since the average surface density measured in this field is likely biased by the presence of the structures ( see 4.2 ) , these are underestimates of the true surface overdensities .
the three groups lie @xmath84 comoving mpc ( i.e. , @xmath85 mpc physical ) apart from one another ( in projection ) , and appear to be connected by filaments traced by the lae distribution .
the three spectroscopically confirmed galaxies at @xmath86 ( shown as blue crosses without red circles ) near the s group may trace a filament running along line of sight .
there are also regions devoid of any lae candidates ( in the northern and central regions and the southeast corner ) .
these areas have comparable photometric depth to the regions in which galaxy overdensities are found , and are not significantly affected by the object mask ( see right panel of figure [ distribution ] ) .
hence , the observed inhomogeneity of the galaxy distribution is likely real .
five of the six highest line - luminosity sources discussed in
[ lae_selection ] ( the large black circles in figure [ distribution ] , left ) lie within or near the three overdense groups : two in the ne group ; two in the s group ; and one just north of the s group , where there is a concentration of four laes .
only one of the six resides far from the galaxy overdensities northeast of the s group .
the distribution of lbg candidates also appears highly inhomogeneous .
a large number of lbgs are observed at the southern edge of the field near ( but not coincident with ) the s group . within the 24@xmath876 area (
dashed grey line in figure [ distribution ] left ) , there are 83 lbgs , 50% larger in number than that expected from the observed field average ( i.e. , @xmath88 ; see [ lbg_selection ] ) . to better visualize and quantify the angular distribution of laes and lbgs , we create two - dimensional density maps of the field .
we first apply a gaussian kernel of fwhm=10 mpc ( i.e. , characteristic radius of 4.25 mpc ) independently to the positional map of the 65 lae candidates and that of the 428 lbg candidates ( see figure [ structure_map ] ) .
the contours show the relative local density with respect to the field .
the discussion of the lae field density estimate is given in
[ overdensity_estimate ] . for the lbgs
, we assume the observed lbg surface density as the field average .
the locations of the three high - overdensity regions , discussed in [ galaxy_distribution ] and shown in figure [ distribution ] , are apparent in the lae density map .
each group resides in a region with the surface density at the core @xmath89 the field average .
the s group is the most significant overdensity , containing 9 laes and 3 additional spectroscopic members .
a smaller , compact group of laes lies @xmath906 ( @xmath9012.6 comoving mpc ; 2.6 physical mpc ) to the north .
we varied the smoothing scales to test the robustness of the overdensity regions . while the overall morphology of the lae density map changes slightly depending on the smoothing scale ,
all three groups are recovered with smoothing scales between @xmath91 mpc . despite the broader redshift distribution resulting from the lbg selection criteria ,
the lbg density map shows a striking resemblance to that derived for the laes .
the two regions of highest lbg density are located in or near the s and ne lae groups .
the lae and lbg distributions associated with the s group are slightly offset from each other , with the lbg surface density forming an elongated structure extending nearly 3 mpc ( physical ) away from the peak of the lae surface density .
the nw group of laes also resides in a region where the lbg density is @xmath92 that of the `` field '' .
the density contour around the s group of lbgs extends northward , mirroring the distribution of laes .
the morphology of lbg density in the region between the nw and s group is also similar to that observed in the lae map .
these similarities lend support to the idea that these are coherent structures in the galaxy distribution .
one notable difference between the lbg and lae maps is the presence of a large population of lbgs in the southeast corner of the field ( @xmath93 the field ) ; no laes are observed there .
differences of this nature might suggest that these are regions where the lbg redshifts ( and the dominant structure ) do not fall within the lae redshift selection window , but are either foreground or background structures .
the line - of - sight distances sampled by lbg and lae candidates are vastly different : the front - to - back comoving distance of the lae sample is @xmath94 mpc , just @xmath95% of that for the lbg sample , i.e. , 480 ( 640 ) mpc , assuming @xmath96 around @xmath76 @xcite .
nevertheless , the spatial density distributions of the two candidate populations resemble each other , and do not appear to be affected by any spatially varying selection criteria .
the close resemblance of the two distribution lends strong support to the fact that many of the lbg candidates must be part of the large - scale structure traced by the laes , and that the discovered structure genuinely represents highly significant matter density peaks traced by both laes and lbgs . in the section that follows ,
we attempt to quantify the overdensities and predict their present - day descendants . in hierarchical galaxy formation theory ,
cluster formation takes place in the most highly overdense regions .
well before their final assembly and subsequent virialization , protocluster regions can be identified via their overall enhancement in space and/or surface density of galaxies ( e.g. , * ? ? ?
* ; * ? ? ?
predicting the total mass overdensity , @xmath97 , from the observed galaxy overdensity , @xmath98 is uncertain without full knowledge of the star- and galaxy - formation physics within dark matter halos ; nevertheless , the observed galaxy overdensity can be used ( in a relative sense ) to gauge the significance of a given structure in the cosmological context and to predict the final mass of its descendant in the present - day universe @xcite .
here , we quantify the significance of the observed galaxy overdensity and begin by estimating the expected number of laes in field environments .
we estimate the expected number of laes selected by our survey by integrating the ly@xmath8 luminosity function derived by @xcite for a @xmath99 lae sample . assuming a schechter function form , @xmath100 $ ] ,
we used @xmath101 , @xmath102 , and @xmath103 , weighted by the filter throughput and the completeness function which defines the depth of our imaging data . the expected number of laes in a given wrc4 magnitude bin ( @xmath104 ) and redshift bin ( @xmath105 ) is given by : @xmath106 where @xmath107 is the ly@xmath8 luminosity function ; @xmath108 is the redshift selection function , defined from the effective filter transmission ( which includes the atmosphere , telescope , and instrument ) , @xmath109 , as @xmath110/max(t_{wrc4})$ ] ; @xmath111 is the comoving volume corresponding to the redshift bin @xmath105 ; @xmath112 is the completeness limit for the wrc4-band ; and where the luminosity function integral is performed within the luminosity limits corresponding to the magnitude bin @xmath104 .
then , the total number of laes is @xmath113 .
based on this analyses , we find that the expected number of field laes in our field is @xmath114 . in comparison ,
the observed number in our field is 65 .
the entire 0.317 deg@xmath12 field is therefore overdense by @xmath115 . using the correct field density
, we can update our earlier estimate , discussed in
[ galaxy_distribution ] , of galaxy overdensity measured within a 8 mpc radius circle as @xmath116 .
if the galaxies trace the underlying mass distribution , what are the likely present - day descendants of these protocluster candidates ?
we can estimate the descendant masses using two approaches , one based on a comparison to numerical simulations , and the other based on a more heuristic approach .
recently , @xcite used the millennium run simulations @xcite coupled with semi - analytic galaxy formation models to predict the expected ( observable ) overdensities associated with protoclusters of a given mass at @xmath16 .
@xcite compute galaxy overdensities using a cubic tophat of volume ( 15 mpc)@xmath3 . in order to compare our observations to their measurements
, we measure the overdensity in a circular area 13 mpc in diameter ; since the line - of - sight depth of our survey is 25 mpc , the total volume enclosed in our overdensity kernel is comparable ( i.e. , @xmath117 )
. the galaxy overdensity is measured on the position of the three groups as @xmath118 where @xmath119 is 1.73 , the mean number of laes within a 13 mpc diameter circle based on the observed lae surface density .
the overdensity will be larger if we use the field density derived in the previous section .
however , we continue to use the observed value in order to derive the most conservative estimates for the masses .
the associated errors are estimated as a quadratic sum of poissonian uncertainties ( @xmath120% ) and cosmic variance of 52% ( see appendix for calculation of cell ( cosmic ) variance ) .
the s , nw and ne groups contain 9 , 8 , and 7 galaxies , respectively , and thus the overdensities in the s , nw and ne groups are [email protected] , [email protected] , [email protected] , respectively .
@xcite find that galaxy overdensities of @xmath121 , 3.6 , and 3.1 observed at @xmath122 will each evolve into @xmath9 galaxy clusters with total mass @xmath123 , @xmath124 , and @xmath125 , respectively ( see their figures 6 , 10 , and associated text ) .
an alternative mass estimate can be made following the approach outlined by @xcite .
the mass associated with a galaxy overdensity that will eventually be bound and virialized by @xmath9 is given by : @xmath126 where @xmath127 is the average matter density of the universe ( @xmath128 ) , @xmath97 is matter overdensity , and @xmath129 is the volume enclosing the observed galaxy overdensity . with the adopted cosmology , equation [ mass_estimate ]
is equivalent to @xmath130~ ( 1+\delta_m)~ [ v/(1~{\rm mpc})^3]$ ] .
the matter overdensity @xmath97 is related to the galaxy overdensity by @xmath131 , where @xmath132 is the correction factor that accounts for the effect of redshift - space distortions @xcite . in the case of spherical collapse
, @xmath132 can be expressed as @xmath133 $ ] . within reasonable ranges of bias @xmath134 and @xmath135 ( @xmath136 , @xmath137 ) ,
the parameter @xmath132 ranges over @xmath138 .
we adopt a bias value for the lae population of @xmath139 ( e.g. , * ? ? ?
* see also the appendix of this paper ) .
we estimate the volume @xmath129 to be the product of the effective transverse area containing the galaxy overdensities and the depth 25 mpc .
the area is estimated from the iso - density contour map ( figure [ structure_map ] ) .
for the ne and nw groups , the transverse area can be approximated as a 9 mpc and 8 mpc radius circle , containing 12 and 8 laes , respectively .
the s group is more elongated as the density contour extends north towards additional 4 laes .
hence we approximate the area associated with the s group as a 10.3 [email protected] mpc rectangular region , containing 12 laes . with the assumptions above
, the total mass associated with each of the observed galaxy overdensities is @xmath140 , @xmath141 , and @xmath142 for the ne , nw , and s group , respectively .
these values are higher than or close to the upper end of the mass estimate according to the @xcite calibration .
the main reason is because the chiang et al .
calibration assumed the galaxy bias of @xmath143 while we used the value @xmath144 , directly measured from the data ( see appendix ) .
adopting their bias value returns the total masses closer to the chiang et al .
estimates ; i.e. , @xmath145 , @xmath141 , and @xmath140 for the ne , nw , and s group , respectively .
galaxy bias ( hence , conversion from galaxy overdensity to matter overdensity ) remains the largest uncertainty in our mass estimate . in summary
, we estimate that the three galaxy overdensities have each @xmath146 measured within a @xmath147 volume .
* their table 4 ) estimate that the regions with @xmath148 have @xmath149% probability to evolve into a galaxy cluster by @xmath9 .
our mass estimates suggest that the nw group will evolve into a fornax - like cluster ( @xmath150 ) , while the two more significant groups will evolve into virgo - like clusters ( @xmath151 ) .
the mass of the s group is nearly @xmath152 ; if the overdensity extends further south , it is possible that its final mass may be even larger comparable to that of the coma cluster ( k .- s .
lee , et al .
, in preparation ) .
we now consider the probability of finding _ three _ overdensities so close to one another .
this is related to the expected number density of a single protocluster in our volume . according to @xcite ,
the number of protoclusters identified to have @xmath153 at @xmath9 is 2,832 in a comoving box of 500 mpc @xmath154 on a side .
hence , the expected number within our survey volume ( @xmath155 ) is almost exactly one .
we expect that the number of galaxy clusters in the wmap7 cosmology , as assumed here , is slightly larger . ]
( i.e. , @xmath156 ) .
an independent estimate from the mass function of x - ray clusters gives a similar answer for the comoving space density @xcite .
given that the expected number is only one , what is the significance of an overdensity of 3 within a @xmath157 volume ?
we estimate this by randomly sampling the millennium runs volume ( assuming the wmap7 cosmology ) on scales of our survey box ( y .- k .
chiang , in private communication ) . of @xmath158 trials within the @xmath159 volume
, we find 2,800 cases in which three or more protoclusters are included in the box .
hence , the probability of finding three or more protoclusters _ of any mass _ within our volume is @xmath160% .
if we restrict the results to only groups where the total mass in protoclusters within the survey box is at least @xmath161 , the probability decreases to 1.9% .
similarly , the probability of finding three or more protoclusters within our survey volume where the most massive protocluster has a total mass of @xmath162 is 1.7 ( 1.2 , 0.9)% . in short , the chances of finding three overdensities matching our observational constraints within the survey volume are very low , @xmath163% .
we would only expect to make _ one _ such discovery if we carried out the lae narrow - band survey to the same depth over @xmath164 deg@xmath12 !
this structure may be one of the largest overdensities , both in size and density , observed thus far .
although the protocluster candidates will likely virialize by @xmath9 , the filamentary structures are unlikely to do so .
these filamentary regions will likely continue to evolve , perhaps becoming highly biased structures similar to local superclusters .
the separation between the three overdensities is comparable to the size of a supercluster in its longest dimension ( see , e.g. , * ? ? ?
most extreme cases of such structures are known as `` great walls '' in the local universe @xcite . to further elaborate on this hypothesis
, we need to firmly establish the physical association between different groups , which will require both larger areal coverage and kinematic information from spectroscopy . finally , we checked the literature to find similarly massive structures containing multiple galaxy overdensities in the volume similar to our field .
most notably , @xcite searched the 2 deg@xmath12 cosmos field using the photometric redshift technique , and identified 36 protocluster candidates at @xmath165 ( see their table 1 ) .
while we do not find any structure containing three protoclusters , there are @xmath166 cases where two candidates are within @xmath167 deg and @xmath168 from each other .
these structures lie at @xmath169 , and typically have @xmath170 .
the relatively low value for the observed overdensity is mainly due to the ` dilution ' caused by the photometric redshift uncertainty of typically @xmath171 in identifying these structures ( equivalent to @xmath172 larger line - of - sight distance than our lae survey ) . based on their simulations covering @xmath173 the cosmos volume ,
@xcite estimated that the probability of any of these candidate protoclusters to evolve into a coma - like cluster ( @xmath174 ) is low , @xmath175 .
based on the number of protocluster pairs found by @xcite , it is possible to independently estimate the likelihood of finding more than one protocluster within our survey volume . over the 1.62 deg@xmath12 area , they found four pairs with @xmath176 deg separation and @xmath177 .
the comoving volume between @xmath178 and @xmath179 is 0.0196 gpc@xmath3 .
therefore , the comoving number density of protocluster pairs is 126 gpc@xmath180 ( @xmath181 ) .
since our survey volume is @xmath2 mpc@xmath182 gpc@xmath3 , the expected number of a protocluster pair within our survey volume is 0.016 .
the estimated likelihood of @xmath183% is likely a conservative limit because the botes structure contains three protoclusters ; no triplet is found in the cosmos protocluster sample . while more detailed comparison between the botes structure and those found in the cosmos field is not possible ( chiang et al .
did not publish mass estimates for their candidates ) , it is assuring that the probability of finding multiple protoclusters in our survey volume is low at just a few percent based on both simulations and real data .
these considerations strongly suggest that the botes structure represents one of the rarest overdensity regions with few known counterparts in the high - redshift universe .
in this paper , we report the discovery of a large - scale structure consisting of three massive galaxy overdensities within a @xmath184 comoving volume at @xmath0 .
deep imaging of the region taken with narrow- and broad - band filters reveal that the number of lyman - alpha emitting galaxies ( or laes ) is @xmath185% higher than that expected in field environments .
the laes in the field show a highly unusual angular distribution ( figure [ distribution ] and [ structure_map ] ) ; three separate regions are observed with a surface overdensity , @xmath186 , of @xmath187 or higher within a 8 mpc radius circle .
these galaxy groups are separated about @xmath188 mpc , and appear to be connected by filaments along which more laes reside .
the distribution of lbgs in the same field resembles that of laes , even though the lbgs sample a factor of 20 - 25 larger line - of - sight distance than laes ; this suggests that these overdensities are indeed massive structures containing both laes and lbgs .
the most significant overdensity lies at the extreme southern end of our image , suggesting that the true extent of this structure may extend beyond the current data . among 65 laes
, we find six sources with high line - luminosity ( @xmath62 , implying @xmath189 ) .
all but one of these high line - luminosity sources reside within or in the vicinity of one of the three galaxy overdensities .
none of the 6 are detected at x - ray wavelengths , and the two that have been observed spectroscopically show no rest - frame uv signatures of agn .
hence , our data suggest that ly@xmath8 emission in these sources may be powered by massive star formation rather than by agn .
recent galaxy simulations demonstrate that at the level of the observed overdensities , these structures will likely be virialized by @xmath9 and become galaxy clusters with masses @xmath190 @xcite .
we estimate that one of the three galaxy groups will have the total mass of @xmath191 and the other two groups will grow in the range of @xmath192
. if the galaxy bias is lower than that assumed in the simulations , the final masses may be even larger . a direct estimate based on the sizes of the overdensities and measured galaxy bias in the region
returns that the total masses enclosed within these regions may be as large as @xmath193 for the two smaller groups , and @xmath194 for the largest one .
given the relatively large distances between these groups ( @xmath195 mpc ) , it is unlikely that all three groups will merge into a single virialized structure by @xmath9 .
nevertheless , it is rare to find three highly significant overdensities so close to one another .
the millennium simulation ( volume @xmath196 billion mpc@xmath3 ) suggests that the probability of chance alignment of three ( or more ) protoclusters within our survey volume ( @xmath197 ) is @xmath1982% ( based on our conservative mass limits given in [ overdensity_estimate ] ) . an independent estimate based on the existing samples of protocluster candidates at high redshift returned a consistently low likelihood .
hence , it appears that we have discovered one of the densest regions of the high - redshift universe with scales well beyond @xmath199 mpc in its longest dimension , similar to the value observed for some of the local superclusters .
further insight into the true extent and kinematic state of this structure , which will help us better assess its ultimate future , will require wider - field imaging data and spectroscopic followup .
identifying a statistical sample of very large structures and their constituents will provide a unique opportunity to study the early evolution of the universe
. we will be able to study the early stages of galaxy formation in different local environments ; from voids , to filaments , to the highest overdensity regions .
measurements of the space density of the protoclusters , and the galaxy distribution within these structures will directly test the theory of structure formation .
unfortunately , such large galaxy overdensities are extremely rare at high redshift and thus are unlikely to be found in large number from deep , small - area surveys . targeted searches over much wider fields ( @xmath200 deg@xmath12 ) will be key to identifying a statistical sample .
ongoing wide - field imaging surveys such as dark energy survey , or future surveys with the large synoptic survey telescope , may provide a sound basis for selecting candidate protoclusters to be targeted .
ksl thanks yi - kuan chiang and roderik overzier for providing the millennium runs data for comparison .
this paper presents data obtained at the kitt peak national observatory .
this research draws upon data provided by [ survey pi ] as distributed by the noao science archive .
noao is operated by the association of universities for research in astronomy ( aura ) under cooperative agreement with the national science foundation .
we are grateful to the expert assistance of the staff of kitt peak national observatory where the optical observations of the ndwfs botes field were obtained .
the authors thank noao for supporting the noao deep wide - field survey .
ad s research was supported in part by the national optical astronomy observatory ( noao ) and by the radcliffe institute for advanced study and the institute for theory and computation at harvard university .
noao is operated by the association of universities for research in astronomy ( aura ) , inc . under a cooperative agreement with the national science foundation . , j. p. , franx , m. , postman , m. , rosati , p. , holden , b. p. , illingworth , g. d. , ford , h. c. , cross , n. j. g. , gronwall , c. , bentez , n. , bouwens , r. j. , broadhurst , t. j. , clampin , m. , demarco , r. , golimowski , d. a. , hartig , g. f. , infante , l. , martel , a. r. , miley , g. k. , menanteau , f. , meurer , g. r. , sirianni , m. , & white , r. l. 2003 , , 596 , l143 , m. , stanford , s. a. , gonzalez , a. h. , zeimann , g. r. , snyder , g. f. , mancone , c. l. , pope , a. , eisenhardt , p. r. , stern , d. , alberts , s. , ashby , m. l. n. , brown , m. j. i. , chary , r .-
r . , dey , a. , galametz , a. , gettings , d. p. , jannuzi , b. t. , miller , e. d. , moustakas , j. , & moustakas , l. a. 2013 , , 779 , 138 , e. , francke , h. , lai , k. , schawinski , k. , gronwall , c. , ciardullo , r. , quadri , r. , orsi , a. , barrientos , l. f. , blanc , g. a. , fazio , g. , feldmeier , j. j. , huang , j .- s . ,
infante , l. , lira , p. , padilla , n. , taylor , e. n. , treister , e. , urry , c. m. , van dokkum , p. g. , & virani , s. n. 2007 , , 671 , 278 , a. , murray , s. s. , forman , w. r. , jones , c. , green , p. , kochanek , c. s. , vikhlinin , a. , fabricant , d. , fazio , g. , brand , k. , brown , m. j. i. , dey , a. , jannuzi , b. t. , najita , j. , mcnamara , b. , shields , j. , & rieke , m. 2005 , , 161 , 9 , c. s. , eisenstein , d. j. , cool , r. j. , caldwell , n. , assef , r. j. , jannuzi , b. t. , jones , c. , murray , s. s. , forman , w. r. , dey , a. , brown , m. j. i. , eisenhardt , p. , gonzalez , a. h. , green , p. , & stern , d. 2012 , , 200 , 8 , e. , smith , k. m. , dunkley , j. , bennett , c. l. , gold , b. , hinshaw , g. , jarosik , n. , larson , d. , nolta , m. r. , page , l. , spergel , d. n. , halpern , m. , hill , r. s. , kogut , a. , limon , m. , meyer , s. s. , odegard , n. , tucker , g. s. , weiland , j. l. , wollack , e. , & wright , e. l. 2011 , , 192 , 18 , g. k. , overzier , r. a. , tsvetanov , z. i. , bouwens , r. j. , bentez , n. , blakeslee , j. p. , ford , h. c. , illingworth , g. d. , postman , m. , rosati , p. , clampin , m. , hartig , g. f. , zirm , a. w. , rttgering , h. j. a. , venemans , b. p. , ardila , d. r. , bartko , f. , broadhurst , t. j. , brown , r. a. , burrows , c. j. , cheng , e. s. , cross , n. j. g. , de breuck , c. , feldman , p. d. , franx , m. , golimowski , d. a. , gronwall , c. , infante , l. , martel , a. r. , menanteau , f. , meurer , g. r. , sirianni , m. , kimble , r. a. , krist , j. e. , sparks , w. b. , tran , h. d. , white , r. l. , & zheng , w. 2004 , , 427 , 47 , m. , shimasaku , k. , akiyama , m. , simpson , c. , saito , t. , ueda , y. , furusawa , h. , sekiguchi , k. , yamada , t. , kodama , t. , kashikawa , n. , okamura , s. , iye , m. , takata , t. , yoshida , m. , & yoshida , m. 2008 , , 176 , 301 , c. , momcheva , i. , willmer , c. n. a. , finkelstein , k. d. , finkelstein , s. l. , tran , k .- v . , brodwin , m. , dunlop , j. s. , farrah , d. , khan , s. a. , lotz , j. , mccarthy , p. , mclure , r. j. , rieke , m. , rudnick , g. , sivanandam , s. , pacaud , f. , & pierre , m. 2010 , , 716 , 1503 , v. , white , s. d. m. , jenkins , a. , frenk , c. s. , yoshida , n. , gao , l. , navarro , j. , thacker , r. , croton , d. , helly , j. , peacock , j. a. , cole , s. , thomas , p. , couchman , h. , evrard , a. , colberg , j. , & pearce , f. 2005 , , 435 , 629 galaxies are biased tracers of underlying matter distribution and estimates of the bias factor are necessary for converting observed galaxy overdensity @xmath135 to the underlying matter overdensity @xmath97 . as a path to estimating the bias , we describe a procedure to measure `` cell variance '' .
cell variance can be easily predicted for any representative galaxy population ( or dark matter ) at any redshift and provides a powerful tool to characterize the large - scale environments in direct comparison with other known galaxy samples .
we define the relative cell variance @xmath201 as : @xmath202 where the first term is the total rms fluctuation in the number of galaxies found in a cell , and the second term accounts for the same quantity arising from the poisson shot noise .
ideally , we want to measure cell variance in a spherical volume . without redshift information of individual sources ,
it is not possible to pinpoint the lae positions in the line - of - sight direction . because the evolution of clustering along the line - of - sight in our volume is negligible , the cell variance measured in a cylindrical volume should be as good as that measured in a sphere of the same volume .
we measure cell variance by counting galaxies within a circle with radius 10 mpc at random positions in our field .
the volume sampled by the 10 mpc circle is the same as that of a sphere of radius @xmath203 mpc ( = 11.4 mpc in the adopted cosmology ) .
we correct the observed counts to account for the area lost to masked regions as @xmath204 , where @xmath205 is the fraction of the total area that is masked out ; we exclude measurements where the loss is more than 60% as unreliable . when averaged over the corrected number counts made over 1000 random trials , we recover the field - averaged surface density within 3% .
we measure @xmath206 for our sample .
the power spectrum normalization for the adopted cosmology is @xmath207 , i.e. , the relative fluctuation of matter at @xmath9 within a @xmath208 radius sphere . in the linear regime , density perturbations grow by a factor of 3.64 from @xmath49 to @xmath9 .
( see equation 19 of @xcite ) ; i.e. , matter fluctuation is @xmath209 .
hence , we estimate that galaxy bias for our laes is @xmath210 ( @xmath211/0.22 $ ] ) .
our value is consistent with the bias estimated from clustering measurements of field laes @xcite , but is slightly lower than the predictions from galaxy simulations ( e.g. , * ? ? ?
* ) derived from the millennium runs , @xmath212 , for all galaxies satisfying the condition @xmath213 . | we report the discovery of a large - scale structure containing multiple protoclusters at @xmath0 in the botes field .
the spectroscopic discovery of five galaxies at @xmath1 lying within 1 mpc of one another led us to undertake a deep narrow- and broad - band imaging survey of the surrounding field . within a comoving volume of @xmath2 mpc@xmath3 ,
we have identified 65 lyman alpha emitter ( lae ) candidates at @xmath4 , and four additional galaxies at @xmath5 .
the galaxy distribution within the field is highly non - uniform , exhibiting three large ( @xmath6 ) overdensities separated by @xmath7 mpc ( physical ) and possibly connected by filamentary structures traced by laes .
the observed number of laes in the entire field is nearly twice the average expected in field environments , based on estimates of the ly@xmath8 luminosity function at these redshifts .
we estimate that by @xmath9 the largest overdensity will grow into a cluster of mass @xmath10 ; the two smaller overdensities will grow into clusters of mass @xmath11 .
the highest concentration of galaxies is located at the southern end of the image , suggesting that the current imaging may not map the true extent of the large scale structure .
finding three large protocluster candidates within a single 0.3 deg@xmath12 field is highly unusual ; expectations from theory suggest that such alignments should occur less than 2% of the time . searching for and characterizing such structures and accurately measuring
their volume space density can therefore place constraints on the theory of structure formation .
such regions can also serve as laboratories for the study of galaxy formation in dense environments . |
@xcite ( also 30 dor b or snr 0538691 ) is a supernova remnant located near the edge of the nebula @xcite in the ( lmc ) .
the complex is the nearest extragalactic giant region and an active star forming region @xcite .
the nebula and the snr are in the same interstellar region of the lmc as and other snrs , among which is , a composite remnant with a pulsar and a plerion component @xcite .
the snr nature of was first determined by its excess of radio over optical emission @xcite , and confirmed by the detection of [ ] lines in the optical spectrum @xcite .
the crab nebula represents the class of pleronic supernova remnants ; it contains a pulsar as a central energy source and a well organised magnetic field indicated by strong linear polarization at radio wavelengths and relatively flat radio spectrum @xcite .
a number of detected snrs have radio characteristics similar to the , yet the pulsar stimulation of the synchrotron nebula is confirmed for just a few of them @xcite . is classified as a crab type remnant because of its centre filled morphology , flat radio spectrum @xcite , and strong soft x
ray emission @xcite . recently a 16ms pulsar detected only in x
rays @xcite has been associated with .
previous observations of at radio wavelength have not detected any polarized emission @xcite .
however , our current investigation at a shorter wavelength shows a significant level of polarization .
all of the data together confirm the classification of as a crab type snr .
the presence of in a relatively dense and complex interstellar environment together with the younger sn1987a and the presumably older snr 0540693 , offers a unique opportunity to study the evolution of crab type snrs in such areas .
the best galactic examples , the crab and 3c58 , both appear to be in exceptionally low
density environments @xcite .
the first obervations at 6 cm were carried out with the australia telescope compact array ( atca ) in the period january
july 1993 .
additional observations at 3.5 and 6 cm were obtained during may 1998 june 1999 .
the six antennas were used in several different array configurations giving baselines with the range from 31 m to 6 km to provide good sampling of the observed region .
details of the atca can be found in @xcite .
the relevant observing parameters are given in table [ tbl - obs ] .
llll date .................................. & 1998 aug 27 & 1993 jan 16 , 19 , 29 & 1998 may 1 , 31 , aug 27 + & 1998 sep 26 , oct 2 & 1993 jul 23 & 1998 sep 26 , oct 222 + & 1999 jun 24 & & 1999 jun 24 + configuations ..................... & 6c , a , 1.5d , 375 & 750b , c , a , d & 750a , e , 6c , a , 1.5d , 375 + field center ( j2000 ) .......... & @xmath2 & @xmath3 & @xmath4 + & @xmath5 & @xmath6 & @xmath5 + total bandwith ( mhz ) ...... & 128 & 128 & 128 + no . of frequency channels
.. & 32 & 32 & 32 + fwhm of primary beam ... & 10 & 10 & 5 + the primary calibrator used in all observations was pks b1934638 with assumed flux densities of @xmath06 and 2.78jy at 6 and 3.5 cm respectively .
pks b0454810 was used as secondary calibration source .
the observations in 1998 1999 were done in mosaic mode with 20 pointing centres to cover the whole region .
the observing pattern was such that the pointing nearest to the snr ( field no.15 see table[tbl - obs ] ) was observed after every five fields to improve signal to noise on this object .
the data were taken simultaneously in 32@xmath74mhz frequency channels of which the 26 central channels are used in the imaging procedure .
they have been averaged down to 13@xmath78mhz channels by hanning smoothing and then to a single channel during imaging .
the data were gridded and deconvolved according to the procedure for mosaic data sets @xcite .
separate images of field no.15 alone were also produced .
the 3.5cm images of the snr were made by combining 7 fields ( including no.15 ) which had the pointing centre closest to the snr .
the deconvolution of all 7 fields was done together , which improves the deconvolution process but introduces some errors in the model of the primary beam .
the edges of the mosaiced region have low sensitivity so the noise level across the mosaic image will vary . in the imaging process the primary beam attenuation will not be completely corrected , because that correction is constrained by the noise in the individual pointings , which can differ @xcite . in this case
, the procedure limits the dynamic range of the final image to 100:1 .
we also made an image of field no.15 alone .
this was cleaned and corrected for the pattern of the primary beam .
the peak brightness in both images is very close ( @xmath07mjy / beam ) , but the single field image has lower signal to noise ratio over the entire field .
the image at 6 cm was constructed by combining four data sets at 4.8ghz and pointing no.15 from 4.4ghz data ( see table [ tbl - obs ] ) .
thus , in addition to the mosaicing procedure explained above , the multi
frequency synthesis technique @xcite was also used to produce the final image of the region . as in the previous case ,
an image of field no.15 alone was made as well .
the interferometer produces an instrumental polarization which increases outward from the center of the primary beam pattern of the individual telescopes . in fields with the pointing centre away from the snr for which an unpolarized source ( 30 doradus ) is at the half
power point , we found fractional polarizations up to @xmath010% with electric field vectors radially oriented with respect to the direction of the beam centre .
this value of fractional polarization in the surrounding fields confirms that off
centre instrumental polarization is indeed strong .
therefore , for the polarimetry we used only field no.15 for which is in the center of the beam and the instrumental polarization is low ( 0.3% ) .
the images of the @xmath8 and @xmath9 stokes parameters were made in a similar way as the @xmath10 image and combined to produce the image of polarized intensity .
to create such an image we used the miriad task impol which statistically corrects the polarized intensity for ricean bias @xcite .
noise in both the @xmath8 and @xmath9 polarization images was below the 5@xmath11 noise level of the total intensity image and cleaning was not necessary .
the polarized intensity images of the snr at 6 cm were made using only field no.15 with the same procedure as for the single
field image explained above .
the 1993 positions were not directly toward n157b and so were not included in the polarization analysis because of the off axis effects disscussed above . the fractional polarization in the direction of was found to be @xmath03% , near the instrumental value , thus confirming only that is not highly polarized at this wavelength .
the high resolution total intensity image at 3.5 cm of the region around was made using all six available configurations ; it has a gaussian beam of hpbw @xmath02 ( see fig.[fig - snr3 cm ] ) .
it shows reasonably uniform brightness distribution and the center filled morphology characteristic of crab type snrs .
we note that the bright peak is resolved .
it lies about @xmath12 ( 4pc at the 50kpc distance to the lmc ) to the northwest of the x
ray peak , which is presumably the location of the pulsar @xcite .
also , the radio peak is not centered on the full snr which extends well to the south with dimensions of about @xmath13 or 30pc@xmath718pc .
slices along the major axis of the remnant at a position angle of about 150@xmath14 show continuous emission throughout the total extent with a slight decrease near a declination of @xmath15 and then a rise again to the south .
perhaps part of an outer shell may be forming around the plerion .
the size given includes this outer emission .
this is the most sensitive high resolution radio image of this object now available , yet no direct evidence of the pulsar can be seen .
this is not surprising , however , as 3.5 cm is a very short wavelength for pulsar emission which usually peaks at meter wavelengths .
the pulsar has not been found at any radio wavelength .
we found strong linear polarization in the direction towards the snr .
the electric field vectors are superimposed on continuum contours in fig.[fig - pol3 cm ] .
the mean fractional polarization is @xmath010% , calculated by dividing the polarized intensity image of the source by the total intensity image ( both made with field no.15 only ) .
the total intensity values were truncated at the 3@xmath11 level which is @xmath05mjybeam@xmath16 before division , but all ( bias corrected ) polarized intensities within the source boundaries given by the total intensity cutoff were used . in the vector display , the vectors were truncated at the 3@xmath11 level .
we note that a polarization map using seven mosaic positions looked similar but was noisier because of the few percent radial instrumental polarization toward from the surrounding fields .
the electric vector alignment is fairly uniform but the polarized intensity is surprisingly patchy .
the highest fractional polarization is 15.4% towards @xmath17 and @xmath18 , whereas at the position of the brightest total intensity @xmath19 and @xmath20 it is 5% , which is the lowest fractional polarization ( excluding blanked regions ) . at the position of the x
ray peak at @xmath21 and @xmath22 the polarized intensity is 7% . the total intensity image of at 6 cm ( shown in fig.[fig - snr6 cm ] )
was made using eight available configurations of the compact array ( see table[tbl - obs ] ) and also has the synthesized beam of hpbw @xmath02 .
the fractional polarization towards ( calculated using only field no.15 ) is only @xmath03% , which is close to the noise level .
we therefore detect no reliable polarization at 6 cm and conclude that there must be strong faraday depolarization between 3.5 and 6 cm .
we note that at 6 cm the distribution of the closest antenna spacings represents lower spatial frequencies than those seen at 3.5 cm .
if the polarization is all on a fine spatial scale we may find a lower value of the polarization at 6 cm .
inspection of fig.[fig - pol3 cm ] , however , shows that while there is fine structure a significant part of the 3.5cm polarization is uniform on scales of over @xmath23 and so should appear at 6 cm .
depolarization effects are obviously present .
is very close to and it is also embedded in an region around the ob association @xcite .
@xcite concluded that the southern part of the snr is obscured and possibly interacts with the dark cloud within the region . with our high radio sensitivity we can see that the snr does extend well to the south of the bright northern component . with a size of 30pc@xmath718pc
, it is the largest crab type snr known .
the feature at @xmath24 and @xmath25 and the elongated structure centered at about @xmath2 and @xmath26 might indicate the beginning of a transition into a composite remnant like snr 0540693 .
the dense interstellar environment of the region may be conducive to the formation of a shell and the irregular structure of may be accounted for by a complex density distribution .
@xcite found that contributes up to 80% of the flux density in this region at 0.843 ghz , the rest coming from the surrounding region .
they suggested that the flat radio spectral index ( @xmath27 ) of @xmath10.1 ( @xmath28 ) indicated from earlier measurements @xcite should be characteristic of the snr itself .
however , they also found that the optical emission lines from originate in conditions associated with blast waves and they therefore concluded that the corresponding non thermal radio emission should not have the flat spectrum typically associated with pulsar stimulation .
our results below differ somewhat from the previously published values derived from low resolution data but still show that n157b has a flat radio spectrum .
data used for determining the radio spectrum of are listed in table [ tbl - spind ] . to obtain a consistent background and source extent , the images were all convolved to the @xmath29 resolution of the most image and the integrated flux density calculated over the same area in each convolved map .
a background intensity around the source was determined from the average of several areas around the snr but far enough from 30 doradus to avoid contamination .
0.843 & @xmath30 & anne green ( 1997 , private comunication ) + 1.400 & @xmath31 & data of @xcite + 2.378 & @xmath32 & data of @xcite + 4.740 & @xmath33 & data of this paper + 8.600 & @xmath34 & data of this paper + the resultant integrated flux densities listed in table [ tbl - spind ] give a mean spectral index for of @xmath10.19 ( see fig.[fig - spind ] ) .
brightness plots also show that the spectrum does not vary significantly across the whole extent of the snr .
this argues that the entire snr is pulsar powered . in the x ray band , however ,
the spectrum of differs slightly relative to other such remnants .
it has a power law energy slope of @xmath11.5 which is steeper than those of other crab - like remnants with values of about @xmath11.0 @xcite .
the x ray and entire spectrum of also strikingly resemble that of snr 0540693 , leaving no doubt of the pleronic origin of .
the snr 0540693 contains a pulsar and is a composite remnant , sharing the characteristics of pure shell remnants and pure plerions @xcite .
this latter composite snr is often called a crab type remnant , because the core is crab like and the faint shell component was only investigated later .
this further similarity between and 0540693 is more evidence that is beginning the transition to a composite remnant .
the polarization properties of the region were previously investigated by @xcite at longer wavelengths .
they used atca observations at 13 cm but concluded that no polarized emission was detected in any region of above the level of instrumental polarization ( less than 2% ) .
we report a significant level of polarized emission at 3.5 cm from for the first time .
this relatively high level of 10% is typical for a crab type remnant .
the data at 6 cm give a marginal detection which we consider an upper limit of 3% fractional polarization at the peak of .
this low level of polarized emission at this wavelength implies very strong internal depolarization .
the observed polarization could also be reduced by the thermal emission from the associated region around which may be intertwined with the snr so that its thermal electrons may cause some of the faraday depolarization as well .
where @xmath36 is the fractional polarization at @xmath37 , the longer wavelength , and @xmath38 , the shorter wavelength ; @xmath39 is the rotation in position angle of the measured electric vector between the two wavelengths .
we will adopt the possible detection of 3% polarization at 6 cm and so use a rough mean value for @xmath40 of 0.3 which requires an average rotation of 135@xmath14 . then using the standard formula for faraday rotation we can estimate the line of sight magnetic field strength : where @xmath39 is in radians , @xmath42 is in meters , @xmath43 is the electron density in @xmath44 , @xmath45 is the magnetic field strength in gauss , and @xmath46 is the path length through the snr in pc . because the faraday rotation increases progressively along the path , we must actually find the adopted rotation along half the depth along the line of sight .
we do not , of course , know this depth directly but choose its value to be approximately equal to the short observed axis of the snr or a half depth of 10pc . from an analysis of rosat and asca
x ray spectra , @xcite derived an electron density of @xmath47@xmath44 , where @xmath48 is the filling factor of the object .
because crab type snrs should be reasonably uniform inside , we adopt a value for @xmath48 of about 0.5 to get a mean density of about 2@xmath44 .
these numbers give a line of sight magnetic field strength of 30@xmath49gauss .
most crab type snrs have rather uniform magnetic fields ; over the limited region where it can be observed , shows more variation than most .
it may be that the field is directed mainly toward us , which would reduce the observed polarization at all wavelengths and also give a large faraday rotation .
no correction has been made for an unknown field orientation which would tend to increase the field strength in the remnant .
we should emphasize that this result is only an order of magnitude estimate .
neglected factors include the uncertainties and probably variations in the depolarization , magnetic field reversals , density uncertainties and variations .
for example , significant clumping can change the apparent integrated electron density because the x ray emission depends upon the density squared but the faraday rotation depends linearly on the density and the sum of squares does not equal the square of sums . where @xmath51 is the total energy in relativistic electrons , @xmath52 is the surface brightness of the remnant at a given observing frequncy @xmath53
, @xmath54 is its distance , and @xmath55 is a parameter that depends on the frequency , the spectral index and the adopted cutoff frequencies for the synchrotron radiation .
there appears to be a break in the spectrum of the non thermal emission between the radio and x
ray @xcite but we do not know specifically where it occurs .
we shall choose an upper frequency of @xmath56hz and a lower one of @xmath57hz , with a spectral index of @xmath10.19 , @xmath58 at 8.6ghz where the flux density is 1.82jy .
with the sizes and magnetic field strength given above , we find that the mean relativistic electron energy density is @xmath59ergs@xmath44 .
this is about 10 times the magnetic energy density .
while within the errors , we can not definitely state that the values are different .
it is interesting to note that in the three objects for which we have now measured the relative energies by this method , each has between 3 and 10 times greater relativistic electron energy than magnetic energy ( kepler s snr @xcite and n23 snr @xcite ) .
@xcite give an approximate temperature for the thermal gas in of 0.4 0.7kev .
choosing 0.6kev and the approximate density of 2@xmath44 used above , the thermal energy density is about @xmath60erg@xmath44 .
thus the thermal energy is about 10 times that in relativistic electrons and 100 times that in the magnetic field . even in this crab type snr
the thermal energy dominates . in summary ,
the filled structure and uniform spectral index of are crab like but it also has some characteristics similar to those of the composite snr 0540693 . although the dense , irregular surroundings may be partly responsible for the non uniform brightness and outline of this snr particularly to the south , the overall morphology and spectral index suggest that this snr is just beginning the transition from a pure crab like remnant to a composite remnant
. it will be interesting to see how it changes at radio , optical and x ray wavelengths in the future .
we have detected for the first time a significant level of linearly polarized emission in this object at 3.5 cm wavelength .
the encompassing h ii region and clumpy internal structure within the snr may cause sufficient faraday rotation to reduce the polarization to below detectable levels at 6 cm and make the large variations in the polarization observed at 3.5 cm at different positions around the remnant .
observations at higher frequencies , e.g. 23ghz , are needed to enable more accurate determination of the faraday rotation .
we thank m. costa , d. milne and h. dickel for participating in the observing .
we thank l. staveley - smith for the atca image at 21 cm and helpful comments .
we thank b. gaensler and a. green for providing us with the most image .
jrd acknowledges support from the campus honours program of the university of illinois at urbana - champaign .
jsl acknowledges support from the astronomy drg and the uws nepean research office .
work at uws nepean was also supported by the arc .
jsl thanks the atnf for the use of their facilities , j. whiteoak for useful comments on polarization , n. killeen and r. sault for valuable discussions on data reduction , and m. filipovic .
bode , j. , 1801 , ` algemeine beschriebung und nachweisung der gestirne ' , berlin : beym verfasser burn , b. j. , 1966 , , 133 , 67 chu , y .- h .
, kennicutt jr . ,
r. c. , schommer , r. a. & laff , j. , 1992 , , 103 , 1545 clarke , j. n. , little , a. g. & mills , b. y. , 1976 , aust .
suppl . , 40 , 1 danziger , i. j. , goss , w. m. , murdin , p. , clark , d. h. & boksenberg , a. , 1981 , , 195 , 33 dickel , j. r. , milne , d. k. , kennicutt , r. c. , chu , y .- h . & schommer , r. a. , 1994 , , 107 , 1067 dickel , j. r. , milne , d. k. , 1998 , , 115 , 1057 fesen , r. a. 1997 , presented in ` 10@xmath61 ergs , ' snr workshop , minneapolis mn , 1997 frater , r. h. , brooks , j. w. & whiteoak , j. b. , 1992 , j. of electrical & electronic eng .
, 12 , 103 ginzburg , v. l. & syrovatskii , s. i. , 1965 , , 3 , 297 gotthelf , e. v. & wang , q. d. , 1996 , ` roentgen strahlung from the universe , ' mpe report 263 , 255 green , a. , 1997 , private communication green , d. a. , 1996 , ` a catalogue of galactic supernova remnants ' , cambridge : mullard radio astronomy observatory , 1996 august version haynes , r. f. , klein , u. , wayte , s. r. , wielebinski , r. , murray , j. d. , bajaja , e. , meinert , d. , buczillowski , u. r. , harnett , j. i. , hunt , a. j. , wark , r. & sciacca , l. , 1991 , , 252 , 475 henize , k. g. , 1956 , , 2 , 315 le marne , a. e. , 1960 , , 139 , 461 long , k. s. & helfand , d. j. , 1979 , , 234 , l77 lucke , p. b. & hodge , p. w. , 1970 , , 75 , 171 manchester , r. n. , staveley - smith , l. & kesteven , m. j. , 1993 , , 411 , 756 marshall , f. e. , gotthelf , e. v. , zhang , w. , middleditch , j. & wang , q.d . , 1998 , , 499 , l179 matsui , y. , long , k. s. , dickel , j. r. & greisen , e. w. , 1984 , , 287 , 295 mebold , u. , deusterberg , c. , dickey , j. m. , staveley - smith , l. & kalberla , p. , 1997
, , 490 , 65 milne , d. k. , caswell , j. l. & haynes , r. f. , 1980 , , 191 , 496 mills , b. y. , turtle , a. j. & watkinson , a. , 1978 , 185 , 263 sault , r. j. & killeen , n. , 1997 , ` miriad users guide , ' atnf publication seward , f. d. , 1983 , in ` supernova remnants and their x - ray emission ' , iau symp . 101 , ( eds . ) j. danziger & p. gorenstain , ( dordrecht , kluwer ) , 405 seward , f. d. , 1989 , space science reviews , 49 , 385 wang , q. d. & gotthelf , e. v. , 1998 , , 494 , 623 weiler , k. w. , 1983 , in ` supernova remnants and their x - ray emission ' , iau symp . 101 , ( eds . ) j. danziger & p. gorenstain , ( dordrecht , kluwer ) , p. 299
weiler , k. w. & sramek , r. a. , 1988 , , 26 , 295 | a new investigation of the supernova remnant ( snr ) was carried out with the australia telescope compact array .
radio continuum images of the entire region have been made at 3.5 and 6 cm wavelength with a resolution of @xmath0 2 .
these data allow a high resolution study of the spectral index distribution and polarization properties of both n157b and the nearby nebula ( the latter will be reported in a subsequent paper ) .
n157b is an extended crab type snr which may be beginning the transition to a composite remnant .
there is little apparent fine structure and the brightest radio region is several parsecs from the probable position of the x ray pulsar .
the snr has a radio spectral index of @xmath10.19 and is significantly polarized at 3.5 cm but not at longer wavelengths . |
in cosmology we attempt to draw large conclusions from limited and often ambiguous data .
i am impressed at how well the enterprise is succeeding , to the point that we have an established standard model for the hot expanding universe ( peebles _ et al .
_ 1991 ) . which elements to include in the standard model is a matter for ongoing debate , of course .
i am inclined to take a conservative line if only to avoid giving misleading impressions to our colleagues with deconstructionist tendencies .
for example , the adiabatic cold dark matter model for structure formation has been more successful than i expected , and as a result is rightly the model most commonly used in studies of how structure might have formed .
simon white calls this model a paradigm , which i take to mean a pattern many find useful and convenient in their research .
i like this use of the term , provided we agree to distinguish it from a well - established standard model .
i think we can not count the adiabatic cold mark matter paradigm as part of the standard model for cosmology because , as argued here , there is a viable and perhaps even more attractive alternative .
i organize this discussion around the issues of the weight of the evidence on whether galaxies are good tracers of mass , what we are learning from the cosmological tests , and the elements of the standard model for structure formation on the scale of galaxies and larger .
i begin with another question , whether einstein s introduction of the cosmological principle set a good example for research in our field .
the hurried reader will find the main points summarized in 6 .
the tension between caution and adventure in the advance of science is well illustrated by the histories these two principles .
einstein ( 1917 ) introduced modern cosmology with his application of general relativity theory to a universe that is spatially homogeneous on average ( that is , a stationary random process ) .
milne gave the homogeneity assumption its name , einstein s cosmological principle .
it is difficult to find in the published literature evidence that einstein was aware of the observational situation on the distribution of matter .
astronomers had established that we live in a bounded island universe of stars , and some had speculated that the spiral nebulae are other island universes .
de sitter ( 1917 ) was willing to consider the possibility that the nebulae are uniformly distributed in the large - scale mean , and that their mass constitutes einstein s near - homogeneous world matter .
on the other hand , de sitter was well aware that the distribution of the nearby nebulae is decidedly clumpy ; indeed , charlier ( 1922 ) pointed out that it resembles a clustering hierarchy ( what we would now call a fractal ) .
that is , the conservative advice from the astronomical community would have been that the observations do not support einstein s world picture , that he would do well to consider a fractal model instead . but now einstein s cosmological principle is well established and part of the standard model : fluctuations from homogeneity on the scale of the hubble length are less than one part in @xmath0 ( from the isotropy of the x - ray background , and about one part in @xmath1 in the standard relativistic model ; peebles 1993 ) .
this is a magnificent triumph of pure thought !
just as the cosmological principle was introduced by hand to solve a theoretical problem , the violation of mach s principle in asymptotically flat spacetime , the biasing principle was introduced to reconcile the low relative peculiar velocities of the galaxies with the high mass density of the theoretically preferred einstein - de sitter world model ( davis _ et al .
there never has been any serious observational evidence for biasing , but the idea rightly was taken seriously because it is elegant and plausible .
but i do not include biasing in the standard model ; we have no very strong evidence for it and the following three arguments against it .
first , there is no identification of a population of void irregular galaxies , remnants of the assumed suppression of galaxy formation in the voids ( peebles 1989 ) .
the first systematic redshift survey showed that the distributions of low and high luminosity galaxies are strikingly similar ( davis _ et al .
i know of no survey since , in 21-cm , infrared , ultraviolet , or low surface brightness optical , that reveals a void population .
there is a straightforward interpretation : the voids are nearly empty because they contain little mass .
second , the improving suite of cosmological tests listed in the next section suggests the mean mass density is well below the einstein - de sitter value .
if the density is low it means galaxies move slowly because there is not much mass to gravitationally pull them , not because they are biased tracers of the mass . third , the galaxy autocorrelation function at low redshift has a simple form , quite close to the power law @xmath2 , with @xmath3 , over three orders of magnitude in separation , @xmath4 .
carlberg shows in these proceedings that the index @xmath5 is quite close to constant back to redshifts near unity . on the theoretical side
, simon white describes elegant numerical simulations of the adiabatic cdm model . in these simulations
the mass autocorrelation function @xmath6 is not close to a power law , and the slope of @xmath6 increases with increasing time .
the two functions allow us to define a bias parameter , b(r , t ) = ^1/2 .
[ eq : bias ] in the adiabatic cdm model this is a function of separation and time .
one interpretation is galaxies are biased tracers of mass , the bias depending on scale and time . but why should the biased tracer exhibit a striking regularity , in @xmath7 and the three - and four - point functions , that is not a property of the mass that is driving evolution ?
the more straightforward reading is that the regularity in @xmath7 reflects a like regularity in the behaviour of the mass , and that there is a slight flaw in the model . given the enormous step we are taking in analyzing the growth of the structure of the universe it surely would not be surprising to learn that we have not yet got it exactly right .
in the standard friedmann - lematre cosmological model coordinates can be assigned so the mean line element is ds^2 = dt^2 - a(t)^2 .
the mean expansion rate satisfies the equation h^2 = ( aa ) ^2 = 83 g + 3 , which can be approximated as h^2= h_o^2[(1+z)^3 + ( 1+z)^2 + ] .
[ eq : cos_pars ] this defines the fractional contributions to the square of the expansion rate by matter , space curvature , and the cosmological constant ( or a term in the stress - energy tensor that acts like one ) .
the time - dependence assumes pressureless matter and constant @xmath8 .
other notations are in the literature ; one that is becoming popular adds the matter and @xmath8 terms , as in michael turner s contribution to these proceedings . to avoid confusion with the definitions in equation ( [ eq : cos_pars ] ) we might express turner s convention as = + .
this isolates the curvature term , which is useful . and
since the evidence is that @xmath9 is small it certainly helps rescue our theoretical preference for a density parameter equal to unity .
i find it unsatisfying , however : what became of the intense debates we had on biasing and the other systematic errors in the measurement of @xmath9 ?
i can get more excited about the full monty : let ^ = + + = 1 - .
[ eq : fm ] each of the terms on the right - hand side of this equation is measurable in principle , and if the applications of the cosmological tests continue to improve at the present rate it may not be many more years before we have ten percent measurements of the three numbers .
if they add to unity we will have a test of general relativity theory applied on large scales in the strong curvature limit .
_ here s my problem : the conference wants latex , the table is set in plain tex because it s much too fiddly for latex , and lanl wo nt accept a postscript file of the compiled file of the table . what is a computer - challenged person to do ? _ the point is illustrated another way in table 1 .
the lines represent quite different ways to probe the standard relativistic model , and the columns are grades for how well three sets of parameter choices fit the results .
as the observations improve we may find that only one narrow range of parameters is consistent with all the constraints .
if so we will have settled two issues .
first , it surely will continue to be difficult to use internal evidence to rule out systematic errors in astronomical observations .
for example , can astronomers unambiguously demonstrate that sneia in a given class of light curve shape really are drawn from the identical population at redshifts @xmath10 and @xmath11 ?
a consistent story from independent tests is strong evidence the measurements have not been corrupted by some subtle systematic error .
second , a consistent story will be a strong positive test of the standard relativistic cosmological model , as in equation ( [ eq : fm ] ) .
the successful parameter set could be quite different from any of the choices in the table , of course ; we may be driven to a dynamical @xmath8 , for example ( peebles & ratra 1988 ; huey _ et al .
_ 1998 ) .
the classical cosmological tests based on measures of the spacetime geometry have been supplemented by a new class of tests based on the condition that the cosmology admit a consistent and observationally acceptable model for structure formation ( categories 3 and 4 in table 1 ) .
i comment on some aspects of structure formation in 4 and 5 .
the constraint from the rate of lensing of quasars by foreground galaxies does not comfortably fit the curvature of the redshift - angular distance relation .
the analysis of lensing by falco _
( 1998 ) , for a combined sample of lensing events detected in the optical and radio , indicates that if the universe is cosmologically flat then @xmath12 at one standard deviation , and @xmath13 at @xmath14 .
the sneia redshift magnitude relation , from the magnificent work of perlmutter _ et al . _
( 1998 ) and reiss _ et al . _
( 1998 ) , seems best fit by @xmath15 , @xmath16 .
the discrepancy is not far outside the error flags , but i think that if the lensing rate were the only available cosmological test we would greet it as confirmation of the einstein - de sitter model and another success for pure thought .
the lensing constraint depends on the galaxy mass function .
the predicted peak of the lensing rate at angular separation @xmath17 arc sec is dominated by the high surface density branch of early - type galaxies at luminosities @xmath18 .
the number density of these objects is not well known , and an improved measurement is an important goal for the new generations of surveys of galaxies . if further tests of the lensing and redshift - magnitude constraints confirm an inconsistency for constant @xmath8 the lesson may be that the cosmological constant is dynamical , rolling to zero , as ratra & quillen ( 1992 ) point out .
the einstein - de sitter model is not yet ruled out , but i think most of us would agree that consideration of structure formation in low density cosmological models is well motivated .
we have good reason to think galaxies grew by gravity out of small initial departures from homogeneity , but the nature of the initial conditions is open to discussion . to illustrate this i present some elements of an isocurvature model .
details are in peebles ( 1998@xmath19 , @xmath20 ) . in
the paradigm simon white describes in these proceedings structure grows out of an adiabatic departure from homogeneity as would be produced by local reversible expansion or contraction from exact homogeneity that is a spatially stationary isotropic random gaussian process .
another possibility is that the primeval mass distribution is exactly homogeneous there is no perturbation to spacetime curvature and structure formation is seeded by an inhomogeneous composition . in the isocurvature model presented here
the initial entropy per baryon is homogeneous , to preserve the paradigm for element formation , and homogeneity is broken by the distribution of cold dark matter . in both models the present mass of the universe is dominated by nonbaryonic cold dark matter ( cdm ) ; i shall call them acdm and icdm models . in the acdm model
the primeval mass density fluctuation ( defined as the most rapidly growing density perturbation mode in time - orthogonal coordinates ) has a close to power law power spectrum , @xmath21 . in the icdm model
the primeval distribution of the cdm is close to a power law , @xmath22 , in a homogeneous net mass distribution .
it is an interesting exercise to check that in linear perturbation theory the evolution from the initial radiation - dominated universe to the present cdm - dominated epoch bends the spectra to @xmath23 where @xmath24 is the wavenumber appearing at the hubble length at the redshift @xmath25 of equality of mass densities in matter and radiation .
the similarity of equations ( [ eq : aspect ] ) and ( [ eq : ispect ] ) for @xmath26 extends to roughly similar spectra of the angular distribution of the thermal cosmic background radiation ( the cbr ) in the adiabatic and isocurvature cdm models .
the status of acdm model fits to the fluctuation spectra of galaxies and the cbr is discussed in these proceedings by bond .
figures 1 and 2 show the icdm model predictions for the parameters m = -1.8,= 0.2 , = 0.8 , h = 0.7 , [ eq : parameters ] with the normalization p(k ) = 6300h^-3 ^ 3 k = 0.1h^-1 , [ eq : pnorm ] where hubble s constant is @xmath27 km s@xmath28 mpc@xmath28 .
the data in figure 1 are from the iras psc - z ( point source catalog ) redshift survey of saunders _ et al . _
this is the real space spectrum after correction for peculiar velocity distortion represented by the density - bias parameter @xmath29 .
there are good measurements of the spectrum of the galaxy distribution on smaller scales , @xmath30 mpc@xmath28 , but this approaches the nonlinear sector , and it seems appropriate to postpone discussion of the small - scale mass distribution until we have analyses of nonlinear evolution from the non - gaussian initial conditions of the model in equation ( [ eq : icdmrho ] ) . since the psc - z catalog is deep , with good sky coverage , it promises to be an excellent probe of the large - scale galaxy distribution , and it is a very useful normalization .
figure 2 shows second moments of the angular distribution of the cbr , where t ( , ) = a_l^m y_l^m ( , ) , t_l = ^1/2 |a_l^m|^2 ^ 1/2 .
[ eq : tl ] in the approximation of the sum over @xmath31 as an integral the variance of the cbr temperature per logarithmic interval of @xmath31 is @xmath32 . in place of @xmath33 , as does bond in his contribution to these proceedings , but since i am considering icdm the convention in equation ( [ eq : tl ] ) , which i prefer because it reflects the @xmath34 components for each value of @xmath31 , may not be unreasonable . ] the measured @xmath35 are from the compilation of ratra ( 1998 ) .
the second moments of the large - scale distributions of mass and radiation in the icdm model agree with the data as well about as could be expected given the state of these difficult measurements .
the same is true of acdm models considered by bond ; both cases pass . at
most one will pass the improved measurements expected from work in progress , but that is for the future .
simple and arguably natural realizations of the inflation concept lead to adiabatic initial conditions ; others to isocurvature initial conditions . in the example of the latter in peebles ( 1998@xmath19 ) the cdm is a scalar field @xmath36 that ends up after inflation in a squeezed state as a gaussian random process with mass density ( * x * ) = m^2(*x*)^2/2 , [ eq : icdmrho ] for field mass @xmath37 . in a simple case
the field satisfies = 0 , ( * x*_1)(*x*_2 ) x_12 ^ - , [ eq : icdme ] and the power spectrum of the mass distribution in equation ( [ eq : icdmrho ] ) is a power law with index @xmath38 .
the model requires @xmath39 , or @xmath40 .
the `` tilt '' from the scale - invariant case @xmath41 is not difficult to arrange ; whether it might be considered natural has yet to be debated .
the primeval density fluctuations in the model in equations ( [ eq : icdmrho ] ) and ( [ eq : icdme ] ) are non - gaussian and scale - invariant : the frequency distribution of the density contrast @xmath42 averaged through a window and scaled by the standard deviation @xmath43 is independent of the window size .
the evidence discussed in peebles ( 1998@xmath20 ) indicates the model with these initial conditions is viable but subject to serious tests from improvements from observational work in progress .
the same is true of the acdm models , of course .
i turn now to one of the tests , the redshift of assembly of the galaxies .
the power law model for the primeval cdm fluctuation spectrum ( eqs .
[ [ eq : ispect ] ] and [ [ eq : parameters ] ] ) is a good approximation for the residual cdm mass distribution at redshifts less than the epoch @xmath25 of equality of mass densities in matter and radiation and on scales small compared to the hubble length at @xmath25 and large compared to the scale of nonlinear clustering . within these bounds the spectrum varies as p_k^md(t)^2 , [ eq : scaling ] where @xmath44 is the solution to the linear equation for the evolution of the density contrast in an isothermal perturbation of the cdm .
the rms contrast through a window of comoving radius @xmath45 varies as x^-(3+m)/2d(t ) .
gravitational structure formation is triggered by passage of upward fluctuations of @xmath42 through unity , and the threshold is not sensitive to @xmath9 in a cosmologically flat model .
this means the characteristic physical length , mass , and internal velocity of newly forming structures scale with time as r_nl(1+z)^-1d^2/(3+m ) , md^6/(3+m ) , ( 1+z)^1/2d^2/(3+m ) .
these relations neglect nongravitational interactions ; they may be expected to be useful approximations on scales much larger than the half - light radii in galaxies , where the cdm halo dominates the mass in the standard model .
we can normalize to the great clusters of galaxies , with r_a & = & 1.5h^-1 , _ cl = 750 ^ -1 , + m_cl & = & 410 ^ 14h^-1 m _ , n_cl = ( 21)10 ^ -6h^-3 ^ -3 .
[ eq : mcl ] the abell radius is @xmath46 , @xmath47 is an rms mean line of sight velocity dispersion for @xmath48 clusters , @xmath49 is the mean mass within the abell radius , and @xmath50 is the present number density of clusters with mass @xmath51 ( bahcall & cen 1993 ) .
clusters are relaxing at the abell radius , and the merging rate is significant , but it is generally agreed that that internal velocities typically are close to what is needed for support against gravity at @xmath52 . in the power law model in equation ( [ eq : scaling ] )
these quantities scaled back in time characterize objects in a like state of early development in the past . with the parameters used in figures 1 and 2 ( eq .
[ [ eq : parameters ] ] ) the scaling relations applied at expansion factor @xmath53 give r_g=15h^-1 , _
g= 140 ^ -1 , m_g=1.310 ^ 11h^-1m_. [ eq : young_galaxies ] the present characteristic separation of clusters and the scaled comoving separation at @xmath53 are d_cl = n_cl^-1/3 = 80h^-1 , d_g = 5h^-1 . in this model
an astronomer sent back in time to @xmath53 would see objects with the somewhat disordered appearance of present - day clusters , merging at a significant rate , but with internal motions typically close to what is needed for virial support .
the characteristic size , mass , and comoving distance between objects would be seen to be characteristic of the luminous parts of present - day @xmath54 galaxies .
our time traveller might well be inclined to call these objects young galaxies , already assembled at @xmath55 . at expansion factor
@xmath56 the scaling relations give r~1,~40 ^ -1 , m~110 ^ 9 m _ ,
numbers characteristic of dwarf galaxies .
i have to assume many merge to form the @xmath54 giants , and that the merging rate eases off at @xmath57 , perhaps because the dissipative settling of the baryons has progressed far enough to lower the cross section for merging , so later structure formation can build the present - day galaxy clustering hierarchy .
if galaxies were assembled as mass concentrations at @xmath53 , as this model suggests , how would they appear at @xmath58 ?
internal velocities ought to be characteristic of present - day galaxies .
that is not inconsistent with the properties of the damped lyman-@xmath59 absorbers studied by wolfe & prochaska ( 1998 ) , though haehnelt , steinmentz & rauch ( 1998 ) show other interpretations are possible .
the expected optical appearance depends on how feedback affects the rate of conversion of gas to stars , a delicate issue i am informed . in these proceedings steidel
presents elegant optical observations of high redshift galaxies that reveal strong spatial clustering .
steidel points out this could signify strong biasing at formation
. the interpretation could be slightly different in the non - gaussian icdm model , where high density fluctuations tend to appear in concentrations ( peebles 1998@xmath60 ) .
structure formation happens later in acdm .
i have expressed doubts that late assembly could produce the high density contrasts of normal present - day @xmath54 galaxies , but the numerical simulations white describes seem not to find this a problem .
if dense galaxies can be assembled at low redshift , when the mean mass density is low , one might have thought that protogalaxies assembled at high redshift and high mean mass density would be unacceptably dense .
but nature was able to form clusters of galaxies that are close to virial equilibrium at modest density contrast at the abell radius , and well enough isolated that they seem likely to remain part of the clustering hierarchy rather than merging into larger monolithic superclusters . under the scaling argument
the same would be true of protogalaxies assembled at @xmath57 in the icdm model .
i arrived at the isocurvature model in 4 ( and peebles 1998@xmath19 and @xmath20 ) through a search for a model for galaxy formation at high redshift , when the cosmic mean density is comparable to that of the luminous parts of a normal large galaxy .
the argument traces back to partridge & peebles ( 1967 ) , a recent version is in peebles ( 1998@xmath60 ) , and elements are reviewed here .
the solid line in figure 3 is the distance between the andromeda nebula m31 and our milky way galaxy in a numerical solution for the motions of the galaxies in and near the local group ( peebles 1996 ) .
the orbits are constrained to arrive at the present positions at expansion parameter @xmath61 from initial motions at @xmath62 consistent with the homogeneous background cosmological model .
this uses the einstein - de sitter model , so the solution may be scaled with time .
the dashed line in the figure assumes spherical symmetry with no orbit crossing , expansion from cosmological initial conditions , and collapse to half the maximum radius , at which point the kinetic energy in the spherical model has reached half the magnitude of the gravitational potential energy .
the solid line for the motion of m31 relative to the milky way has a similar shape but with significant differences . in the numerical solution neighbouring galaxies are close to the milky way and m31 at @xmath63
, so the solid line is more strongly curved than a spherical solution with fixed mass .
the solid line is less strongly curved at larger expansion factor because the interaction with neighbouring mass concentrations has given the milky way and m31 substantial relative angular momentum .
the present transverse relative velocity of m31 is comparable to the radial velocity of approach , the minimum separation is about half the present value , and the mean separation in the future is larger than the present value . as we all know , nonradial motions tend to suppress collapse .
now let us consider the spherical solution as a model for young galaxies .
let @xmath64 be the proper radius of a sphere that is centred on the young galaxy and contains the mass @xmath65 in equation ( [ eq : young_galaxies ] ) . in the spherical solution , which ignores nonradial motion and the motion of mass across the surface of the sphere ,
the radius varies with time as @xmath66 , where @xmath67 and @xmath68 .
this ignores the cosmological constant @xmath8 , which has little effect on the orbit .
if spherical collapse stops at radius @xmath69 at redshift @xmath70 , then in the spherical model the collapse factor from maximum expansion is r_g / r_max = ( 1 - _ g)/2 , [ eq : cf ] and an adequate approximation to @xmath71 is = 89(_g h_or_g ) ^2(1+z_g)^-3 = 410 ^ 4(1+z_g)^3 , [ eq : etao ] for the numbers in equation ( [ eq : young_galaxies ] )
. for the collapse factor @xmath72 in the dashed line in figure 3 equation ( [ eq : etao ] ) says @xmath73 , not far from the value @xmath74 in equation ( [ eq : young_galaxies ] ) . in a model for late galaxy assembly , at @xmath75
, equations ( [ eq : cf ] ) and ( [ eq : etao ] ) say @xmath76 .
this is the dotted line in figure 3 .
the pronounced collapse could result from exchange of energy among lumps settling out of a more extended system , as happens in numerical simulations ( navarro , frenk , & white 1996 ) , but i think there are two reasons to doubt it happens in galaxy formation .
first , in a hierarchical model for structure formation collapse to @xmath77 kpc at @xmath78 traces back to a cloud of subgalaxy fragments
star clusters at radius @xmath79 kpc .
i know of no evidence of such clustering ( apart from the usual power law correlation functions ) in deep samples .
second , the scaled process of formation of rich clusters of galaxies shows no evidence of pronounced collapse : clusters seem to be close to stable at the abell radius and present in significant numbers at redshift @xmath80 .
these are arguments , not demonstrations .
i consider them persuasive enough to lend support to the isocurvature model that leads from the fit to measures of large - scale structure in figures 1 and 2 to the scaling model for early galaxy assembly in equation ( [ eq : young_galaxies ] ) .
this depends on whether galaxies really were assembled early , of course , and are fortunate that the observations steidel describes in there proceedings may well be capable of telling us when the galaxies formed .
it is inevitable that the exciting rush of advances in this subject has left ideas unexplored .
i have attempted to identify some roads not taken , less popular lines of thought that seem worth considering .
the main points are summarized in the following questions .
einstein s brilliant success in establishing key elements of the standard cosmological model is an example of why we pay serious attention to elegant ideas even in the face of contrary empirical indications .
but i think this is not an entirely edifying example : einstein s intuition was not always so successful , and most of us are not einsteins . in the present still crude state of cosmology
it is better to be led by the phenomenology from astronomy and from particle physics ( that may teach us the identity of the dark matter , for example ) .
most of us agree that the einstein - de sitter model is the elegant case , and it makes sense that the community has given it special attention despite the long - standing indication from galaxy peculiar velocities that the einstein - de sitter density is too high .
now other lines of evidence are pointing in the same direction , as summarized in table 1 , and i think there is general agreement in the community that we must give serious consideration to the possibility that nature has other ideas about elegance .
i count this as a cautionary example for the exploration of ideas on how the galaxies formed .
the galaxy two - point correlation function is quite close to a power law , @xmath81 , over three orders of magnitude of separation @xmath82 at low redshift , and the index @xmath5 is quite close to constant back to redshifts approaching unity .
this is not true of the mass autocorrelation function @xmath6 in the adiabatic cold dark matter ( acdm ) model .
thus we have a measure of bias , @xmath83 ^{1/2}$ ] ( eq . [
[ eq : bias ] ] ) , that depends on position and time .
should we take this as evidence galaxies are biased mass tracers ? since the regularity is in the galaxies surely the first possibility to consider is that @xmath7 is revealing a like regularity in the behaviour of the mass , that the bias is in the model .
this reading is heavily influenced by a related issue : if much of the cdm is in the voids defined by normal galaxies where are the remnants of the void galaxies ?
surely they are not entirely invisible ?
i am impressed by the elegant simulations of the acdm models simon white presents in these proceedings , and have to believe they reflect aspects of reality .
but the curious issue of @xmath84 leads me to suspect there is more to the story .
would the isocurvature variant do better ?
that awaits searching tests by numerical simulations of the kind that that have been applied to the adiabatic case .
one often reads that it is to determine how the world ends .
but should we trust an extrapolation into the indefinitely remote future of a theory that we know can only be a good approximation to reality ? for a trivial example , suppose the universe has zero space curvature and the present value of the density parameter in matter capable of clustering is @xmath15 , with the rest of the contribution to @xmath85 in a term that acts like a cosmological `` constant '' @xmath8 that is rolling toward zero ( peebles & ratra 1988 ; huey _ et al .
if the final value of @xmath8 is identically zero then the world ends as minkowksi spacetime ( after all the black holes have evaporated ) . if @xmath8 ends up at a permanent negative value , no matter how close to zero , the world ends in a big crunch .
should we care which it is ?
i would consider a bare answer an empty advance , because the excitement of physical science is in discovering the interconnections among phenomena .
perhaps the excitement of knowing how the world ends will be in what it teaches us about how the world began .
the classical cosmological tests , that probe spacetime geometry , have been greatly enriched by tests based on the condition that the cosmology admit a consistent and observationally acceptable theory for structure formation .
the structure formation theory in turn tests ideas about what the universe was like before it was well described by the classical friedmann - lematre model , and may eventually allow us to enlarge the standard model to include the story of how the world begins and ends .
generally accepted elements are the gravitational growth of small primeval departures from homogeneity , that may be described as a stationary isotropic random process , in a universe with present mass that is dominated by cdm and maybe a term that acts like a cosmological constant .
the most striking piece of evidence for the gravitational instability picture is the agreement between the primeval density fluctuations needed to produce the cbr anisotropy and the present distribution and motion of the galaxies .
precision measurements in progress should allow us to fix many of the details of this gravitational instability picture , but within present constraints we can not say that the primeval density fluctuations are gaussian , or adiabatic , because we have a viable alternative , the non - gaussian isocurvature model mentioned in 4 .
the main piece of evidence for the cdm is the mismatch between the baryon mass density in the standard model for the origin of the light elements and the mass density indicated by dynamical analyses of relative motions of the galaxies .
our reliance on hypothetical mass is embarrassing ; a laboratory demonstration of its existence would be an exceedingly valuable advance .
it is a sign of the growing maturity of our field that we can pose questions that are motivated by specific theoretical issues and can be addressed by feasible observations .
but i think our subject still is immature enough that we should be quite prepared for surprises .
my favorite example is shaver s ( 1991 ) demonstration that the radio galaxies within @xmath86 mpc distance are close to the plane of the local supercluster , even though the plane is not apparent in the general distribution of galaxies at this depth .
if the clusters and radio sources were produced by a pancake collapse why do we not see it in the general galaxy distribution ?
maybe a better picture is that in the early universe a nearly straight cosmic string passed by , piling mass in its wake into a sheet that fragmented into the seeds of engines of active galaxies .
i think the most surprising outcome of the new surveys would be that there are no major corrections to what we think we know .
this work was supported in part by the usa national science foundation . | i argue that the weight of the available evidence favours the conclusions that galaxies are unbiased tracers of mass , the mean mass density ( excluding a cosmological constant or its equivalent ) is less than the critical einstein - de sitter value , and an isocurvature model for structure formation offers a viable and arguably attractive model for the early assembly of galaxies . if valid these conclusions complicate our work of adding structure formation to the standard model for cosmology , but it seems sensible to pay attention to evidence .
psfig [ firstpage ] |
we report the results of a numerical study of the two - dimensional ( 2d ) members of a new family of quasiperiodic tilings proposed recently by vidal and mosseri @xcite .
these new so - called generalized rauzy tilings ( grt ) ( after the work of @xcite ) , are generalizations of the fibonacci chain and are simpler in structure than hitherto studied tilings in two and three dimensions .
tight - binding models on these systems may be amenable to analytical techniques whereas other 2d tilings thus far studied have been refractory to approaches including renormalization group , or recursive methods @xcite .
this property makes the new tilings attractive candidates to study in the world of quasiperiodic hamiltonians .
it is thus interesting to examine whether the tight - binding spectrum of these new tilings have similarly complex structures as for the canonical " cases studied earlier , namely the penrose or the ammann - beenkker ( octagonal ) tilings .
this turns out not to be the case : these tilings fall in a class intermediate between the periodic lattice , an integrable problem , and the octagonal tiling , which belongs in the non - integrable class .
this conclusion is based on calculations of statistical properties of the energy levels of periodic approximants of these new systems .
such analyses have been carried out for complex systems , including disordered media , systems with interactions or quantum billiards @xcite .
a relation is found to exist between the quantum statistics and the classical dynamics : depending on whether the dynamics is classically ergodic , or classically chaotic , one finds their quantum properties to be described by poisson statistics or by random matrix theory ( rmt ) @xcite .
such statistical analyses are useful for gaining insight into complex systems for which exact , or even approximate theoretical results are not available .
this is the case for the octagonal tiling , a quasiperiodic tiling in two dimensions which has been tackled via numerical study of tight - binding models @xcite .
we show that the grt in two dimensions has spectral properties that differ @xmath1 from those of the octagonal tiling .
the grt appears analogous to the pseudointegrable billiards proposed in @xcite which are neither completely integrable nor completely chaotic .
the cut - and - project method provides a simple and easily implemented way to obtain pieces of quasiperiodic tilings .
one can thus obtain arbitrarily large pieces of an infinite quasicrystal , or one can construct finite pieces that can be periodically continued
these are the periodic approximants of the quasicrystal .
as described in @xcite , one first defines the real - space directions , which constitute a d - dimensional subspace of a larger d - dimensional hyperspace .
the remaining d - d directions constitute the perpendicular " space .
the quasicrystal is obtained by projecting all the points contained within a certain infinite strip in the hyperspace @xcite onto the real space .
when the orientation of the real space is irrational , the tiling is aperiodic , while if rational orientations are chosen , the tiling repeats periodically in real space .
this method has been used to obtain the fibonacci chain ( d=2 , d=1 ) , the octagonal tiling ( d=4 , d=2 ) , the three - dimensional penrose tiling ( d=6 , d=3 ) etc .
the codimension , d - d , is in fact equal to d in the three cited cases .
the simplifying feature of the grt @xcite is that their codimension is equal to 1 for all real space dimension d. for a finite tiling , this allows one to index the points according to their perpendicular space coordinate .
furthermore , when the points of an approximant of these tilings illustrated for two dimensions in fig.1 are so indexed , the connectivity matrix of the approximant ( assuming toroidal boundary conditions ) takes on a very simple band - diagonal ( toeplitz ) structure @xcite . in the two dimensional case that we consider here ,
the nonzero matrix elements are situated at distances of @xmath2 and @xmath3 from the diagonal .
the generalized fibonacci numbers @xmath4 are obtained from a three term recursion relation @xmath5 , with the initial conditions @xmath6 .
the ratio @xmath7 tends to the value @xmath8 , solution of the equation @xmath9 .
these matrices correspond to approximants of increasing size as @xmath10 is increased .
the model we consider is a pure hopping tight - binding hamiltonian , @xmath11 where @xmath12 in the sum correspond to pairs of sites linked by a bond in fig.1 .
the hopping amplitude has been assumed to be independent of the linked sites and is set to unity .
we consider the case of periodic boundary conditions , and take @xmath13 where @xmath14 is the bloch vector .
the resulting eigenvalue problem reduces to that of diagonalizing the connectivity matrix defined above .
this was done using a lanczos routine .
the spectrum is expected to be symmetric for the infinite tiling , since it is bipartite .
in addition , there is a discrete symmetry under an inversion in the perpendicular subspace @xcite .
this results in two subspaces of positive and negative parity respectively , and our analyses are carried out for each sector separately .
the spectral density @xmath15 is shown in fig.2a , with that of the octagonal tiling shown in 2b for comparison ( the localized states of the latter at @xmath16 have not been included in the figure ) .
the density of states curves shown were obtained for given sample sizes assuming periodic boundary conditions , and illustrate the characteristic rapidly fluctuating behavior typical of these aperiodic tilings ( similar curves were found in @xcite and @xcite ) .
it is evident that while the grt has an underlying smoothly varying component of @xmath15 which is everywhere non - zero ( notice the y - axis offset ) whereas the octagonal tiling does not .
calculations on approximants of the 2d penrose tiling show strong fluctuations as well with a multitude of gaps and pseudogaps @xcite .
the envelope of @xmath15 of the grt is in fact reminiscent of the density of states of the square lattice .
one notes in fig.1 that there are many regions in which sites of coordination number 4 are grouped together in this tiling .
in contrast to the octagonal tiling , which also has a mean coordination number @xmath17 , the fluctuations of geometry are less strong in the grt . this is a result of the reduced codimension , and we see that in consequence the density of states has a much less singular structure as compared to the octagonal case .
turning to the statistics of nearest neighbor level spacings , we investigate the distribution of @xmath18 where the @xmath19 are the ordered set of energy levels of the hamiltonian ( 1 ) . in order to eliminate the trivial dependence of this quantity on variations of @xmath15 , one first carries out an unfolding " of the spectrum @xcite to get the corrected level spacings @xcite .
this yields the unfolded spacings @xmath20 whose mean value is equal to 1 .
the resulting distribution of spacings @xmath21 is plotted in fig.3a for the three largest systems studied , which contain 35738(l ) , 19513(m ) and 10609(s ) sites .
the dashed line represents the poisson @xmath22 decay , while the continuous line is the semi - poisson form @xmath23 .
the latter has been proposed for pseudointegrable billiards and can be shown to correspond to a certain short range plasma model ( srpm ) by bogomolny et al @xcite .
these authors speculate that this type of statistics is a consequence of fractal structure of the wave functions in those systems . in the anderson model ,
the critical level statistics were first shown to have the semi - poisson distribution by braun et al @xcite . in another example of this universality class evangelou and pichard
had considered the distribution of bandwidths and obtained the semi - poisson form in the critical harper model @xcite . in fig.3a one
sees that there is a size - dependence in the fall - off in @xmath21 at small @xmath24 . for the largest size ,
the points appear to deviate substantially from the semi - poisson curve
. it will be interesting to calculate the next system size to see if the deviations are systematically upwards , i.e. towards the poisson limit .
the @xmath21 resembles that obtained in @xcite for periodic boundary conditions at the critical point of the anderson model .
in contrast to the random case however , in our case there is a size - dependence .
it is important to note that the @xmath25 of unfolding matters here : nonunfolded spacings follow a simple poisson law while progressively stronger unfolding causes the distribution to go over to the semi - poisson form . in fig.3b
we present a semilog plot of the statistics of the maximally unfolded level spacings ( where we renormalized each spacing by the locally averaged value over the nearest neighbor energy levels ) .
the agreement with the semi - poisson form is seen to be good over a wide range of values . at large @xmath24 @xmath21
falls off faster than exponentially in @xmath24 , while at small @xmath24 , there is a size dependent increase of @xmath21 above the semi - poisson value .
we believe that these effects may be explained by the unfolding procedure , although boundary condition effects can not be ruled out and further studies are needed to settle this issue .
this distribution in fig.3a is very different from the one obtained for the octagonal tiling . for that tiling ,
unfolded spacings follow the wigner - dyson distribution for random matrices of the gaussian orthogonal ensemble ( goe ) , @xmath26 . if one considers the distribution of the bare " spacings ( i.e. without correcting for the variation in density of states ) , they follow a broad ( log - normal ) distribution law : this can be shown via a recursive relation based on the inflation symmetry of the tilings @xcite .
a less singular case is that of a randomized version of the octagonal tiling @xcite .
the randomized tiling has a less fluctuating density of states and the spacings distribution is simply goe ( unfolding does not produce any new effects ) . from these @xmath21 , as well as from the spectral rigidity function ( to be discussed below )
we conclude that the octagonal tilings belong in the class of non - integrable models exemplified by disordered metals .
one has an unbounded diffusion of wave packets resulting in level repulsion at small @xmath24 .
the grt is clearly not in the goe regime for any scale of energies , rather it should be considered a nearly integrable system , as the shape of @xmath21 indicates .
this becomes more evident upon looking at the spectral rigidity function @xmath27 , where @xmath28 is the number of levels lying within an energy interval of width @xmath29 .
the angular brackets mean averages over different starting positions of the interval within the spectrum .
fig.4a shows the plot of @xmath30 for the largest tiling size , for three different choices of unfolding @xcite the continuous line in fig.4a indicates the srpm result @xmath31 .
for very small @xmath29 , the dependence appears well described by the semi - poisson law .
the three curves have the same low energy behavior , but the most unfolded " case has its fluctuations damped out because of the unfolding , while the other cases show the transition from approximately linear to approximately quadratic dependence on @xmath29 . the srpm behavior at small @xmath29 may be related to a long time dynamics in this tiling which is intermediate between chaotic and integrable .
this would correspond to the semi - poisson form of @xmath21 it is interesting to consider the rigidity evaluated for the energy levels without unfolding . in the case of random media , where diagrammatic perturbation theory can be exploited , and checked by more semiclassical methods @xcite , it was shown that in the diffusive regime , as a function of energy @xmath32
, @xmath33 has two different sorts of behavior : firstly , at very low energies , it has the logarithmic dependence expected for goe matrices .
this holds upto the scale of the thouless energy @xmath34 @xcite which is size dependent , corresponding to the time taken to diffuse to the boundaries . for larger energies @xmath35 where the exponent @xmath36 describes the spreading over time of an initially localized wave packet @xmath37
this theoretical prediction was verified numerically in the anderson model @xcite , where the exponent is one for ordinary diffusion , @xmath38 . for the randomized octagonal tiling , @xmath39 @xcite as deduced from @xmath30 , in agreement with the results of a direct calculation of wavepacket diffusion on this tiling @xcite .
fig.4b shows the log - log plot of @xmath30 as a function of energy for three sizes .
there is a crossover from linear dependence to an almost quadratic dependence , @xmath40 ( @xmath41 ) .
the sub - quadratic regime very likely corresponds to the sub - ballistic regime of dynamics that was observed for wavepackets diffusing on a 2d grt by vidal and mosseri @xcite .
the two - dimensional grt tiling , despite its quasiperiodic structure , is an almost integrable system , with a semi - poisson form for the level spacings distribution and a corresponding small - energy spectral rigidity function .
the @xmath21 distribution is probably linked to the existence of fractal wavefunctions , as has already been observed in critical quantum systems . in the grt studied here , it is noteworthy that although fluctuations of the density of states are present , they are strongly reduced with respect to those in canonical " quasiperiodic tilings .
this is a consequence of the reduced codimension of the tilings , which leads to its having weaker quasiperiodic modulations of geometry than the previously studied octagonal and penrose tilings .
further study is needed to elucidate the nature of states and of the relation between dynamical evolution of wavepackets and the calculated statistical properties of the energy levels .
i would like to acknowledge useful discussions with frederic piechon and michel duneau .
i thank the computing center at idris , orsay for the computing facilities and support needed for this study .
mesoscopic quantum physics , proceedings of les houches summer school eds .
e. akkermans et al ( elsevier , 1996 ) ; chaos and quantum physics , proceedings of les houches summer school , eds . m.j .
giannoni , a. voros and j. zinn - justin , ( elsevier , new york 1991 ) m. duneau and a. katz , phys .
lett . * 54 * , 2688 ( 1985 ) ; p. a. kalugin , a. y. kitaev , and l. s. levitov , j. phys .
( paris ) lett . *
46 * , l601 ( 1985 ) ; m.duneau , r.mosseri , and c.oguey , j. phys . a * 22 * , 4549 ( 1989 )
the hamiltonian is invariant under a reindexation of points given by inversion of the perpendicular space coordinate with respect to the midpoint of the strip .
this symmetry is not exactly obeyed in the approximants , however the symmetry breaking gets weaker as system size is increased . | we consider a set of tilings proposed recently as d - dimensional generalisations of the fibonacci chain , by vidal and mosseri .
these tilings have a particularly simple theoretical description , making them appealing candidates for analytical solutions for electronic properties . given their self - similar geometry
, one could expect that the tightbinding spectra of these tilings might possess the characteristically singular features of well known quasiperiodic systems such as the penrose or the octagonal tilings .
we show here , by a numerical study of statistical properties of the tight - binding spectra that these tilings fall rather in an intermediate category between the crystal and the quasicrystal , @xmath0 in a class of almost integrable models .
this is certainly a consequence of the low codimension of the new tilings . |
one of the most challenging and fascinating goals of modern physics is the achievement of a fully quantitative understanding of the strong interaction , which is the subject of hadron physics .
significant progress has been achieved over the past few years thanks to considerable advances in experiment and theory .
new experimental results have stimulated a very intense theoretical activity and a refinement of the theoretical tools .
still there are many fundamental questions which remain basically unanswered .
phenomena such as the confinement of quarks , the existence of glueballs and hybrids , the origin of the masses of hadrons in the context of the breaking of chiral symmetry are long - standing puzzles and represent the intellectual challenge in our attempt to understand the nature of the strong interaction and of hadronic matter .
experimentally , studies of hadron structure can be performed with different probes such electrons , pions , kaons , protons or antiprotons . in antiproton - proton annihilation particles with gluonic degrees of freedom as well as
particle - antiparticle pairs are copiously produced , allowing spectroscopic studies with very high statistics and precision .
therefore , antiprotons are an excellent tool to address the open problems
. the recently approved fair facility ( facility for antiproton and ion research ) , which will be built as a major upgrade of the existing gsi laboratory in germany , will provide antiproton beams of the highest quality in terms of intensity and resolution , which will provide an excellent tool to answer these fundamental questions .
the panda experiment ( pbar annihilations at darmstadt ) will use the antiproton beam from the high - energy storage ring ( hesr ) colliding with an internal proton target and a general purpose spectrometer to carry out a rich and diversified hadron physics program , which includes charmonium and open charm spectroscopy , the search for exotic hadrons and the study of in - medium modifications of hadron masses .
this paper is organized as follows : in section 2 we will give an overview of the fair facility and the hesr ; in section 3 we will discuss some of the most significant items of the panda experimental program ; in section 4 we will give a brief description of the panda detector .
finally in section 5 we will present our conclusions .
the planned fair complex is shown in fig .
the heart of the system consists of two synchrotron rings , called sis100 and sis300 , housed in the same tunnel , which will provide proton and ion beams of unprecedented quality .
the sis100 , a 100 t@xmath0 m proton ring , will feed the radioactive ion and antiproton beam lines for experiments to be carried out in the high - energy storage ring ( hesr ) , the collector and cooler rings ( cr ) and the new experimental storage ring ( nesr ) .
the sis300 will deliver high energy ion beams for the study of ultra relativistic heavy ion collisions .
the accelerators of fair will feature significant improvements in system parameters over existing facilities : * * beam intensity * : increased by a factor of 100 to 1000 for primary and 10000 for secondary beams ; * * beam energy * will increase by a factor 30 for heavy ions ; * * beam variety * : fair will offer a variety of beam lines , from antiprotons to protons , to uranium and radioactive ions ; * * beam precision * : availability of cooled antiproton and ion beams ( stochastic and electron cooling ) ; * * parallel operation * : full accelerator performance for up to four different , independent experiments and experimental programs .
these features will make fair a first rate facility for experiments in particle , nuclear , atomic , plasma and applied physics .
the antiproton beam will be produced by a primary proton beam from the sis100 .
the @xmath1 production rate will be of approximately @xmath2/s .
after @xmath3 have been produced they will be transferred to the hesr , where internal experiments in the @xmath1 momentum range from 1 gev / c to 15 gev / c can be performed .
[ fig : hesr ] the layout of the hesr is shown in fig .
2 . it is a racetrack ring , 574 meters in length , with two straight sections which will host the electron cooling and , respectively , the panda experiment .
two modes of operation are foreseen : in the high - luminosity mode peak luminosities of @xmath4@xmath5s@xmath6 will be reached with a beam momentum spread @xmath7 , achieved by means of stochastic cooling ; in the high - resolution mode for beam momenta below 8 gev / c electron cooling will yield a smaller beam momentum spread @xmath8 at a reduced luminosity of @xmath9@xmath5s@xmath6 .
the high - resolution mode will allow to measure directly the total width of very narrow ( below 1 mev ) resonances .
the panda experiment has a rich experimental program whose ultimate aim is to improve our knowledge of the strong interaction and of hadron structure .
the experiment is being designed to fully exploit the extraordinary physics potential arising from the availability of high - intensity , cooled antiproton beams .
significant progress beyond the present understanding of the field is expected thanks to improvements in statistics and precision of the data .
many experiments are foreseen in panda . in this paper
we will discuss the following : * charmonium spectroscopy ; * search for gluonic excitations ( hybrids and glueballs ) ; * study of hadrons in nuclear matter ; * open charm spectroscopy .
ever since its discovery in 1974 @xcite charmonium has been a powerful tool for the understanding of the strong interaction .
the high mass of the @xmath10 quark ( m@xmath11 @xmath12 1.5 gev / c@xmath13 ) makes it plausible to attempt a description of the dynamical properties of the ( @xmath14 ) system in terms of non - relativistic potential models , in which the functional form of the potential is chosen to reproduce the asymptotic properties of the strong interaction .
the free parameters in these models are to be determined from a comparison with the experimental data .
now , more than thirty years after the @xmath15 discovery , charmonium physics continues to be an exciting and interesting field of research .
the recent discoveries of new states ( @xmath16 , x(3872 ) ) , and the exploitation of the b factories as rich sources of charmonium states have given rise to renewed interest in heavy quarkonia , and stimulated a lot of experimental and theoretical activities .
over the past few years a significant progress has been achieved by lattice gauge theory calculations , which have become increasingly more capable of dealing quantitatively with non perturbative dynamics in all its aspects , starting from the first principles of qcd .
experimentally charmonium has been studied mainly in @xmath17 and @xmath18 experiments . in @xmath17 annihilations direct charmonium formation
is possible only for states with the quantum numbers of the photon @xmath19 , namely the @xmath15 , @xmath20 and @xmath21 resonances .
precise measurements of the masses and widths of these states can be obtained from the energy of the electron and positron beams , which are known with good accuracy .
all other states can be reached by means of other production mechanisms , such as photon - photon fusion , initial state radiation , b - meson decay and double charmonium . on the other hand
all @xmath14 states can be directly formed in @xmath18 annihilations , through the coherent annihilation of the three quarks in the proton with the three antiquarks in the antiproton .
this technique , originally proposed by p. dalpiaz in 1979 @xcite , could be successfully employed a few years later at cern and fermilab thanks to the development of stochastic cooling . with this method
the masses and widths of all charmonium states can be measured with excellent accuracy , determined by the very precise knowledge of the initial @xmath18 state and not limited by the resolution of the detector .
the cross section for the process : @xmath22 is given ( in units @xmath23 ) by the well known breit - wigner formula : @xmath24 where @xmath25 and @xmath26 are the center - of - mass ( c.m . ) energy and momentum ; @xmath27 , @xmath28 and @xmath29 are the resonance spin , mass and total width and @xmath30 and @xmath31 are the branching ratios into the initial ( @xmath18 ) and final states . due to the finite energy spread of the beam ,
the measured cross section is a convolution of the breit - wigner cross section , eq .
( [ eq : bw ] ) , and the beam energy distribution function @xmath32 ; the effective production rate @xmath33 is given by : @xmath34 where @xmath35 is the instanteneous luminosity , @xmath36 an overall efficiency @xmath37 acceptance factor and @xmath38 a background term .
the parameters of a given resonance can be extracted by measuring the formation rate for that resonance as a function of the c.m .
. the accurate determination of masses and widths depends crucially on the precise knowledge of the absolute energy scale and on the beam energy spectrum . carried out at fermilab ( a ) and beam energy distribution in each data point ( b ) . [
fig : chi1scan],width=302 ] the technique is illustrated in fig .
[ fig : chi1scan ] which shows a scan of the @xmath40 resonance carried out at the fermilab antiproton accumulator by the e835 experiment @xcite using the process @xmath41 .
for each point of the scan the horizontal error bar in ( a ) corresponds to the width of the beam energy distribution .
the actual beam energy distribution is shown in ( b ) .
this scan allowed the e835 experiment to carry out the most precise measurement of the mass ( @xmath42 mev / c@xmath13 ) and total width ( @xmath43 mev ) of this resonance . the spectrum of charmonium states is shown in fig . [ fig : ccspectrum ] .
, width=302 ] it consists of eight narrow states below the open charm threshold ( 3.73 gev ) and several tens of states above the threshold .
all eight states * below @xmath44 threshold * are well established , but whereas the triplet states are measured with very good accuracy , the same can not be said for the singlet states .
the @xmath45 was discovered almost thirty years ago and many measurements of its mass and total width exist , with six new measurements in the last four years . still the situation is far from satisfactory .
the particle data group ( pdg ) @xcite value of the mass is @xmath46 mev / c@xmath13 , an average of eight measurements with an internal confidence level of 0.026 : the error on the @xmath45 mass is still as large as 1.2 mev / c@xmath13 , to be compared with few tens of kev / c@xmath13 for the @xmath15 and @xmath20 and few hundreds of kev / c@xmath13 for the @xmath47 states .
the situation is even worse for the total width : the pdg average is @xmath48 mev , with an overall confidence level of only 0.001 and individual measurements ranging from 7 mev to 34.3 mev .
the most recent measurements have shown that the @xmath45 width is larger than was previously believed , with values which are difficult to accomodate in quark models .
this situation points to the need for new high - precision measurements of the @xmath45 parameters .
the first experimental evidence of the @xmath49 was reported by the crystal ball collaboration @xcite , but this finding was not confirmed in subsequent searches in @xmath18 or @xmath17 experiments .
the @xmath49 was finally discovered by the belle collaboration @xcite in the hadronic decay of the b meson @xmath50 with a mass which was incompatible with the crystal ball candidate .
the belle finding was then confirmed by cleo @xcite and babar @xcite which observed this state in two - photon fusion .
the pdg value of the mass is @xmath51 mev / c@xmath13 , corresponding to a surprisingly small hyperfine splitting of @xmath52 mev / c@xmath13 , whereas the total width is only measured with an accuracy of 50% .
the study of this state has just started and all its properties need to be measured with good accuracy .
the @xmath53 state of charmonium ( @xmath54 ) is of particular importance in the determination of the spin - dependent component of the @xmath55 confinement potential .
the fermilab experiment e760 reported an @xmath54 candidate in the decay channel @xmath56 @xcite , with a mass of @xmath57 mev / c@xmath13 .
this finding was not confirmed by the successor experiment e835 , which however observed an enhancement in the @xmath58 @xcite final state at a mass of @xmath59 mev / c@xmath13 .
the @xmath54 was finally observed by the cleo collaboration @xcite in the process @xmath60 with @xmath61 , in which the @xmath45 was identified via its hadronic decays .
they found a value for the mass of @xmath62 mev / c@xmath13 .
it is clear that the study of this state has just started and that many more measurements will be needed to determine its properties , in particular the width .
the region * above @xmath44 threshold * is rich in interesting new physics . in this region , close to the @xmath44 threshold ,
one expects to find the four 1d states .
of these only the @xmath63 , identified with the @xmath64(3770 ) resonance , has been found .
the @xmath65 states ( @xmath66 and @xmath67 ) are predicted to be narrow , because parity conservation forbids their decay to @xmath44 .
in addition to the d states , the radial excitations of the s and p states are predicted to occur above the open charm threshold .
none of these states have been positively identified .
the experimental knowledge of this energy region comes from data taken at the early @xmath17 experiments at slac and desy and , more recently , at the b - factories , cleo - c and bes .
the structures and the higher vector states observed by the early @xmath17 experiments have not all been confirmed by the latest much more accurate measurements by bes @xcite .
a lot of new states have recently been discovered at the b - factories , mainly in the hadronic decays of the meson : these new states ( x , y , z ... ) are associated with charmonium because they decay predominantly into charmonium states such as the @xmath15 or the @xmath20 , but their interpretation is far from obvious .
the situation can be roughly summarized as follows : * the z(3931 ) @xcite , observed in two - photon fusion and decaying predominantly into @xmath44 , is tentatively identified with the @xmath68(2s ) ; * the x(3940 ) @xcite , observed in double charmonium events , is tentatively identified with the @xmath45(3s ) ; * for all other new states ( x(3872 ) , y(3940 ) , y(4260 ) , y(4320 ) and so on ) the interpretation is not at all clear , with speculations ranging from the missing @xmath14 states , to molecules , tetraquark states , and hybrids .
it is obvious that further measurements are needed to determine the nature of these new resonances
. the main challenge of the next years will be thus to understand what these new states are and to match these experimental findings to the theoretical expectations for charmonium above threshold .
charmonium spectroscopy is one of the main items in the experimental program of panda , and the design of the detector and of the accelerator are optimized to be well suited for this kind of physics .
panda will represent a substantial improvement over the fermilab experiments e760 and e835 : * up to ten times higher instantaneous luminosity ( @xmath69@xmath5s@xmath6 in high - luminosity mode , compared to @xmath70@xmath5s@xmath6 at fermilab ) ; * better beam momentum resolution ( @xmath71 in high - resolution mode , compared with @xmath72 at fermilab ) ; * a better detector ( higher angular coverage , magnetic field , ability to detect the hadronic decay modes ) . at full luminosity panda
will be able to collect several thousand @xmath14 states per day . by means of fine scans
it will be possible to measure masses with accuracies of the order of 100 kev and widths to 10% or better .
the entire energy region below and above open charm threshold will be explored .
one of the main challenges of hadron physics , and an important item in the panda physics program , is the search for gluonic excitations , i.e. hadrons in which the gluons can act as principal components . these _ _
g__luonic hadrons fall into two main categories : glueballs , i.e. states of pure glue , and hybrids , which consist of a @xmath55 pair and excited glue .
the additional degrees of freedom carried by gluons allow these hybrids and glueballs to have @xmath73 exotic quantum numbers : in this case mixing effects with nearby @xmath55 states are excluded and this makes their experimental identification easier .
the properties of glueballs and hybrids are determined by the long - distance features of qcd and their study will yield fundamental insight into the structure of the qcd vacuum .
antiproton - proton annihilations provide a very favourable environment in which to look for gluonic hadrons .
two particles , first seen in @xmath74 scattering @xcite with exotic quantum numbers @xmath75 , @xmath76 @xcite and @xmath77 @xcite , are clearly seen in @xmath18 annihilation at rest . on the other hand a narrow state at 1500 mev / c@xmath13 discovered in @xmath18 annihilations by the crystal barrel experiment
@xcite , is considered the best candidate for the glueball ground state ( @xmath78 ) , even though the mixing with nearby @xmath55 states makes this interpretation difficult .
so far the experimental search for glueballs and hybrids has been mainly carried out in the mass region below 2.2 mev / c@xmath13 .
panda will extend the search to higher masses and in particular to the charmonium mass region , were light quark states form a structure - less continuum and heavy quark states are far fewer in number . therefore exotic hadrons in this mass region could be resolved and identified unambiguously .
the spectrum of charmonium hybrid mesons can be calculated within the framework of various theoretical models , such as the bag model , the flux tube model , the constituent quark model and recently , with increasing precision , from lattice qcd ( lqcd ) . for these calculations
the parameters are fixed according to the properties of the known @xmath55 states .
all model predictions and lqcd calculations agree that the masses of the lowest lying charmonium hybrids are between 4.2 gev / c@xmath13 and 4.5 gev / c@xmath13 .
three of these states are expected to have @xmath73 exotic quantum numbers ( @xmath79 , @xmath80 , @xmath81 ) , making their experimental identification easier since they will not mix with nearby @xmath14 states .
these states are expected to be narrower than conventional charmonium , because their decay to open charm will be suppressed or forbidden below the @xmath82 threshold .
the cross sections for the formation and production of charmonium hybrids are estimated to be similar to those of normal charmonium states , which are within experimental reach .
formation experiments will generate only non - exotic charmonium hybrids , whereas production experiments will yield both exotic and non - exotic states .
this feature can be exploited experimentally : the observation of a state in production but not in formation will be , in itself , a strong hint of exotic behavior .
the glueball spectrum can be calculated within the framework of lqcd in the quenched approximation @xcite . in the mass range accessible to panda as many as 15 glueball states
are predicted , some with exotic quantum numbers ( _ oddballs _ ) . as with hybrids ,
exotic glueballs are easier to identify experimentally since they do not mix with conventional mesons .
the complications arising from mixing with normal @xmath55 states is well illustrated by the case of the f@xmath83(1500 ) .
as mentioned above , this narrow state , observed at lear by the crystal barrel @xcite and obelix @xcite experiments , is considered the best candidate for the ground state glueball .
however this interpretation is not unique , and relies on the combined analysis of the complete set of two - body decays of the f@xmath83(1500 ) and two other scalar states , the f@xmath83(1370 ) and the f@xmath83(1710 ) .
this analysis yields the following mixing picture @xcite : @xmath84 .
other scenarios for the scalar meson nonet not involving a glueball have been proposed and this makes the interpretation of the f@xmath83(1500 ) as the ground state glueball ambiguous .
this example highlights the need to extend the glueball search to higher mass regions , which are free of the problem of mixing with conventional @xmath55 states .
the study of medium modifications of hadrons embedded in hadronic matter is aimed at understanding the origin of hadron masses in the context of spontaneous chiral symmetry breaking in qcd and its partial restoration in a hadronic environment .
so far experiments have been focussed on the light quark sector : evidence of mass changes for pions and kaons have been deduced by the study of deeply bound pionic atoms @xcite and of k meson production in proton - nucleus and heavy - ion collisions @xcite .
the high - intensity @xmath1 beam of up to 15 gev / c will allow an extension of this program to the charm sector both for hadrons with hidden and open charm .
the in - medium masses of these states are expected to be affected primarily by the gluon condensate .
recent theoretical calculations predict small mass shifts ( 5 - 10 mev / c@xmath13 ) for the low - lying charmonium states @xcite and more consistent effects for the @xmath47 ( 40 mev / c@xmath13 ) , @xmath20 ( 100 mev / c@xmath13 ) and @xmath64(3770 ) ( 140 mev / c@xmath13 ) @xcite .
d mesons , on the other hand , offer the unique opportunity to study the in - medium dynamics of a system with a single light quark .
recent theoretical calculations agree in the prediction of a mass splitting for d mesons in nuclear matter but , unfortunately , they disagree in sign and size of the effect .
experimentally the in - medium masses of charmonium states can be reconstructed from their decay into di - leptons and photons , which are not affected by final state interaction .
d meson masses , on the other hand , need to be reconstructed by their weak decays into pions and kaons which makes the direct measurement of mass modifications difficult .
therefore other signals have been proposed for the detection of in - medium mass shifts of d mesons : in particular it has been speculated that a lowering of the @xmath44 threshold would result in an increased @xmath85 and @xmath86 production in @xmath1-nucleus annihilations @xcite or in an increase in width of the charmonium states lying close to the threshold @xcite .
another study which can be carried out in panda is the measurement of @xmath15 and d meson production cross sections in @xmath1 annihilation on a series of nuclear targets .
the comparison of the resonant @xmath15 yield obtained from @xmath1 annihilation on protons and different nuclear targets allows to deduce the @xmath15-nucleus dissociation cross section , a fundamental parameter to understand @xmath15 suppression in relativistic heavy ion collisions interpreted as a signal for quark - gluon plasma formation .
the hesr running at full luminosity and at @xmath1 momenta larger than 6.4 gev / c would produce a large number of @xmath85 meson pairs .
the high yield ( e.g. 100 charm pairs per second around the @xmath64(4040 ) ) and the well defined production kinematics of @xmath85 meson pairs would allow to carry out a significant charmed meson spectroscopy program which would include , for example , the rich @xmath85 and @xmath87 meson spectra .
[ dspectrum ] spectrum.,width=302 ] the b - factory experiments have discovered several new resonances in the @xmath85 and @xmath87 sectors , where two are extremely narrow : the @xmath88(2317 ) @xcite and the @xmath88(2317 ) @xcite .
these new states appear at unexpected locations , since their masses are more than 140 mev / c@xmath13 lower than expected from potential models , as shown in fig .
this has given rise to speculations about their nature .
it is important to verify these findings by means of new measurements .
threshold pair production can be employed for precision measurements of the mass and the width of the narrow excited @xmath85 states .
in order to carry out the physics program discussed above the panda detector must fulfil a number of requirements : it must provide ( nearly ) full solid angle coverage , it must be able to handle high rates ( @xmath89 annihilations / s ) with good particle identification and momentum resolution for @xmath90 , e , @xmath91 , @xmath92 , k and p. additional requirements include vertex reconstruction capability and , for charmonium , a pointlike interaction region , efficient lepton identification and excellent calorimetry ( both in terms of resolution and of sensitivity to low - energy showers ) . a schematic view of the panda detector is shown in fig .
[ fig : det ] .
the antiprotons circulating in the hesr hit an internal hydrogen target ( either pellet or cluster jet ) , while for the nuclear part of the experimental program wire or fiber targets will be used .
the apparatus consists of a central detector , called target spectrometer ( ts ) and a forward spectrometer ( fs ) .
the ts , for the measurement of particles emitted at laboratory angles larger than 5@xmath93 , will be located inside a solenoidal magnet which provides a field of 2 t. its main components will be a microvertex silicon detector , a central tracker ( either a straw tube detector or a time projection chamber ) , an inner time - of - flight telescope , a cylindrical dirc ( detector of internally reflected light ) for particle identification , an electromagnetic calorimeter consisting of pbwo@xmath94 crystals , a set of muon counters and of multiwire drift chambers . the fs will detect particles emitted at polar angles below 10@xmath93 in the horizontal and 5@xmath93 in the vertical direction .
it will consist of a 2 t@xmath0 m dipole magnet , with tracking detectors ( straw tubes or multiwire chambers ) before and after for charged particle tracking .
particle identification will be achieved by means of erenkov and time - of - flight detectors .
other components of the fs are and electromagnetic and a hadron calorimeter .
all detector components are currently being developed within a very active r&d program .
this continued development implies that the choice has not yet been finalized for all detector elements .
the availability of high - intensity , cooled antiproton beams at fair will make it possible to perform a very rich experimental program .
* precision measurement of the parameters of all charmonium states , both below and above open charm threshold , with the possible discovery of the missing states ( e.g. the d - wave states ) , which will lead to a full understanding of the charmonium spectrum ; * the observation / discovery of glueballs and hybrids , particularly in the mass range between 3 and 5 gev / c@xmath13 , yielding new insights into the structure of the qcd vacuum ; * the measurement of mass shifts of charmonium and open charm mesons in nuclear matter , related to the partial restoration of qcd chiral symmetry in a dense nuclear medium ; * open charm spectroscopy ( @xmath85 and @xmath87 spectra ) .
all these new measurements will make it possible to achieve a very significant progress in our understanding of qcd and the strong interaction .
we are looking forward to many years of exciting hadron physics at fair .
99 j. j. aubert et al .
* 33*(1974)1404 ; + j. e. augustin et al . , _ phys .
lett . _ * 33*(1974)1406 .
p.dalpiaz , proceedings of the _ _ f__irst lear workshop , karlsruhe , germany , edited by h. poth ( 1979 ) .
m. andreotti et al . , _ nucl .
_ * b717 * ( 2005 ) 34 .
yao et al . , _ j. of phys . _ * g33 * ( 2006 ) 1 .
c. edwards et al .
* 48 * ( 1982 ) 70 .
s. k. choi et al . , _ phys .
* 89 * ( 2002 ) 102001 .
d. m. asner et al .
* 92 * ( 2004 ) 142001 .
b. aubert et al .
, _ phys . rev .
lett . _ * 92 * ( 2004 ) 142002 .
t. a. armstrong et al . , _ phys .
* 69 * ( 1992 ) 2337 .
m. andreotti et al . , _ phys .
* d 72 * ( 2005 ) 032001 .
j. l. rosner et al.,_phys .
* 95 * ( 2005 ) 102003 .
j. z. bai et al . , _ phys .
* 84 * ( 2000 ) 594200 ; + j. z. bai et al .
lett . _ * 88 * ( 2002 ) 101802 .
s. uehara et al .
* 96 * ( 2006 ) 082003 .
k. abe et al . , _ phys .
_ * 94 * ( 2005 ) 182002 .
d. thompson et al . , _ phys .
* 79 * ( 1997 ) 1630 ; + g. adams et al .
_ * 81 * ( 1998 ) 5760 .
a. abele et al .
* b 423 * ( 1998 ) 175 .
j. reinnarth et al . , _ nucl .
phys . _ * a 692 * ( 2001 ) 268c . c. amsler et al . , _ phys .
lett . _ * b 355 * ( 1995 ) 425 .
c. morningstar and m. peardon , _ phys .
rev . _ * d 60*(1999)34509 .
m. bargiotti et al . , _ phys .
* b 561 * ( 2003 ) 233 .
f. e. close and a. kirk , _ phys .
* b483 * ( 2000 ) 345 . h. geissel et al . , _ phys .
* 88 * ( 2002 ) 122301 ; + h. geissel et al .
_ * b549 * ( 2002 ) 64 ; + h. geissel et al .
* 92 * ( 2002 ) 072302 .
m. nekipelov et al . , _ phys . lett .
_ * b540 * ( 2002 ) 207 ; + z. rudy et al . , _ eur .
phys . j. _
* a15 * ( 2002 ) 303 ; + y. shin et al .
_ * 81 * ( 1998 ) 1576 ; + r. barth et al .
* 78 * ( 1997 ) 4007 ; + f. laue et al .
* 82 * ( 1999 ) 1640 ; + p. crochet et al .
_ * b486 * ( 2000 ) 6 ; + k. wisniewski et al .
phys . j. _ * a9 * ( 2000 ) 515 .
w. klingl et al . , _ nucl .
_ * a624 * ( 1997 ) 527
. s. lee and c. ko , _ phys .
_ * c67 * ( 2003 ) 038202 ; + s. lee , proceedings of int . workshop `` hadron 2003 '' , edited by h. k. e. klempt and h. orth . ,
aschaffenburg , 2004 .
a. sibirtsev , k. tsushima and a. thomas , _ eur .
phys . j. _ * a6 * ( 1999 ) 351 .
a. hayashigaki , _ phys .
* b487 * ( 2000 ) 96 ; + b. friman , s. lee and t. song , _ phys .
_ * b548 * ( 2002 ) 153 .
b. aubert et al .
_ * 90 * ( 2003 ) 242001 .
d. besson et al . , _ phys.rev . _
* d68 * ( 2003)032002 ; + p.
krokovny et al . , _ phys.rev.lett . _
* 91 * ( 2003 ) 262002 ; + y. mikami et al . ,
_ phys.rev.lett . _
* 92 * ( 2004 ) 012002 . | the physics program of the future fair facility covers a wide range of topics that address central issues of strong interactions and qcd .
the antiproton beam of unprecedented quality in the momentum range from 1 gev / c to 15 gev / c will allow the panda experiment to make high precision , high statistics measurements , which include charmonium and open charm spectroscopy , the search for exotic hadrons and the study of in - medium modifications of hadron masses . |
platonic solids are understood here as the subset of polytopes whose vertices are generated , starting from a single point in @xmath0 , @xmath9 , by the action of a finite coxeter group .
platonic polytopes are distinguished by the fact that their faces @xmath10 of any dimension @xmath2 are platonic solids of lower dimension , and that they are transformed into each other by the action of the coxeter group , i.e. , they belong to one orbit of the corresponding coxeter group . it has been known since antiquity that there are five platonic solids in @xmath11 @xcite , namely , the regular tetrahedron , cube , octahedron , icosahedron and dodecahedron ( see fig .
[ plato ] ) .
the underlying coxeter groups , @xmath7 , @xmath8 . ] are @xmath12 , @xmath13 , @xmath14 and @xmath7 .
they are the lowest - dimensional cases we consider by our method .
the root systems of finite coxeter groups of any type @xcite allow one to extend our method to higher dimensions . for @xmath15
, a classification of platonic solids was done more than century ago by schlfi @xcite . in this case , the coxeter groups are @xmath16 , @xmath17 , @xmath18 , @xmath6 and @xmath8 . it is also known that , in any dimension @xmath19 , there are only three such solids generated by groups of types @xmath3 , @xmath4 and @xmath5 , namely the simplex , the hypercube and the cross - polytope .
they correspond to the regular tetrahedron , cube and octahedron in 3-dimensional space . for every platonic solid there exists its dual , which is also a platonic solid .
we describe both members of each dual pair . for @xmath9 , the dual pair of @xmath3
consists of two identical solids oriented differently in space .
the platonic solids of @xmath4 and @xmath5 form the dual pair . for @xmath6 , @xmath7 and @xmath8 ,
the dual pairs are formed by different solids . in this paper
, we derive these results by new rather simple means , using elementary decoration rules for the corresponding connected coxeter - dynkin diagrams @xcite . we also provide a constructive method for building faces @xmath10 of dimensions @xmath2 .
although we focus on polytopes of dimension @xmath20 , it is also useful to consider @xmath21 platonic solids because they occur as 2-dimensional faces of higher - dimensional polytopes .
the general idea of the diagram decoration method @xcite is to consider the coxeter group @xmath22 of the simple lie algebra @xmath23 , that is , the symmetry group of a given solid , and to identify the subgroup @xmath24 that pointwise stabilizes the given face @xmath10 , and the subgroup @xmath25 , which is the symmetry group of the face @xmath10 .
then we have @xmath26 since both @xmath24 and @xmath25 are generated by reflections defined by some of the simple roots of @xmath23 , it is possible , indeed convenient , to distinguish by different decoration the simple roots of the coxeter - dynkin diagram defining the reflections generating the two subgroups .
the decorated diagrams allow one to identify all of the polytope faces that belong to different orbits of the corresponding coxeter groups , and to count how many times each face occurs on the polytope .
nodes of the coxeter - dynkin diagram we decorate by one of the three symbols , @xmath27 , @xmath28 correspond respectively to @xmath29 , @xmath27 , @xmath30 , in @xcite and to @xmath31 , @xmath32 , @xmath30 in @xcite . ] , @xmath33 , @xmath27 , @xmath28 , according to the rules in subsection [ rules ] . decorated coxeter - dynkin diagrams are a powerful method of great generality @xcite , which could be used to solve other problems , but which so far remain underused . in @xcite ,
the method was used to describe voronoi and delone cells in root lattices of all simple lie algebras @xmath23 .
it still has to be used to describe the voronoi and delone cells in weight lattices @xcite , and for other problems as well @xcite ... in this paper , we use a version @xcite of the method to describe the platonic solids together with all of their faces . in dimension 4
, equivalent results can be found among the entries of table 3 of @xcite .
decoration rules are recursive . starting from a seed decoration , which provides information about the vertices @xmath34 of the polytope , the procedure consists in modifying the decoration step - by - step . at each step
, the modified decoration provides information about the faces of dimension greater by 1 .
decoration rules are quite general and are particularly simple when applied to platonic polytopes .
the same set of decorations applies to coxeter - dynkin diagrams with the same number of nodes .
as the links between diagram nodes do not affect decoration rules , the links need not be drawn .
only when working with a specific reflection group , the decorated diagram with all links removed is viewed superimposed on the appropriate coxeter - dynkin diagram .
in general , the number of copies of the face @xmath10 , contained in a given polytope , is equal to the ratio of orders of the coxeter groups @xcite , @xmath35 where @xmath36 is the order of the corresponding reflection group .
the subgroups @xmath37 and @xmath38 are read as sub - diagrams of the coxeter - dynkin diagram of @xmath23 .
their nodes are identified by the appropriate decoration for @xmath10 , namely @xmath28 for reflections generating @xmath37 , and @xmath33 for @xmath38 .
the symmetry group of an intersection of two faces , say @xmath39 and @xmath40 of dimensions @xmath41 and @xmath42 , respectively , is the reflection group @xmath43 generated by reflections decorated by black rhombi and appearing in the diagram of either face .
see the motivating example [ motiv ] . or , to answer the inverse of this question : given a face , say @xmath39 , of a polytope , how many faces @xmath10 of higher dimension , @xmath44 , have @xmath39 in common ?
frequently , one would want to know the number of edges originating in the same vertex , @xmath45 , or the number of faces @xmath10 meeting in an edge ( @xmath46 ) .
the answer is the size of the orbit of the stabilizer @xmath47 in @xmath22 when it is acting on @xmath10 .
the stabilizer is the reflection group generated by the reflections labeled by rhombi of both colours in the decorated diagram of @xmath39 .
see examples [ edgmeetver ] , [ facemeet ] .
the well - known results of the classification of finite dimensional simple lie algebras of any rank and type @xmath48 are used here only to identify the finite reflection group @xmath22 , called the weyl group , or , equivalently , the crystallographic coxeter group . in addition .
we also consider polytopes with symmetries of finite non - crystallographic coxeter groups denoted by @xmath49 , @xmath7 , @xmath8 , together with their simple root diagrams .
.number @xmath50 of non - zero roots of simple lie algebras @xmath23 and the orders @xmath36 of their weyl group .
number of roots and the orders of the three non - crystallographic coxeter groups . [ cols="^,^,^,^,^,^,^,^,^,^",options="header " , ] the geometry of the set of simple roots ( relative lengths and relative angles ) in the real euclidean space @xmath0 is described by well - known conventions implied in drawing the corresponding coxeter - dynkin diagrams ( see , for example , @xcite ) .
reflections generating the coxeter groups act in the @xmath51-dimensional real euclidean space @xmath0 spanned by the simple roots @xmath52 , according to @xmath53 that is , @xmath54 is the reflection in the hyperplane of dimension @xmath55 , containing the origin of @xmath0 , and orthogonal to the simple root @xmath56 , @xmath57 .
instead of @xmath58-basis of simple roots , we use @xmath59-basis .
two bases are linked by the cartan matrix @xmath60 , ( for example @xcite ) @xmath61 the decorations described in this paper should be viewed superimposed on the appropriate coxeter - dynkin diagram . often , the same decoration applies to several diagrams .
it should therefore be read in the context of that diagram when some specific information about the face of the polytope needs to be deduced from it .
decoration rules for connected coxeter - dynkin diagrams are recursive .
they apply to diagrams of all simple lie algebras and to any polytope generated by the corresponding coxeter group starting from a single point in @xmath0 .
they can therefore be written without reference to platonic solids . the ` grammar ' for every decoration : * a node of the diagram can be decorated by @xmath27 , @xmath28 , or by @xmath33 . * in a diagram of @xmath51-nodes , there can be up to @xmath51 squares placed in any node . * any connected pair of nodes must not carry @xmath28 and @xmath33 side by side .
assuming that a starting decoration complies with the grammar rules , the decoration describes a face of dimension equal to the number of @xmath33 in it .
for dual polytopes , the same decoration refers to the face of dimensions equal to the number of @xmath28 .
usually one starts from vertices , faces of 0-dimension , in which case there is no @xmath33 .
decoration rules : 1 .
replace one of the squares by @xmath33 .
2 . change to @xmath27 each @xmath28 that became adjacent to the new @xmath33 .
3 . repeat steps ( 1 ) and ( 2 ) as long as there are any @xmath28 .
the following proposition is a direct consequence of the decoration rules .
a polytope has one coxeter group orbit of faces @xmath10 for every dimension @xmath2 , i.e. it is platonic , provided ( i ) the coxeter - dynkin diagram forms a connected line with no branches or loops .
( ii ) the seed decoration has one square placed at either of the extreme nodes of the coxeter - dynkin diagram .
[ coxdiag ] coxeter - dynkin diagrams that give rise to platonic solids in @xmath0 , where @xmath62 , are of types in this paper , we define dual polytopes and their faces using decorated coxeter - dynkin diagrams : the role of decoration elements , @xmath28 and @xmath33 , are reversed in the dual polytope .
that is , @xmath28 specifies reflections that generate the symmetry group @xmath38 of the dual face , while @xmath33 provides the generating reflections of the subgroup @xmath37 that stabilizes the face pointwise .
the dual polytope of a platonic polytope is also platonic .
specifically , we have dual pairs of platonic polytopes in @xmath0 , @xmath9 , @xmath63 a particular decoration that carries information about a face @xmath10 of a polytope in @xmath0 can be read as information about the face of the dual polytope , say @xmath64 .
let us illustrate the decoration method on the transparent example before presenting the general rules . consider the classical platonic solids in @xmath11 .
connected coxeter - dynkin diagrams of the groups , generated by three reflections , are of types @xmath12 , @xmath13 , @xmath14 and @xmath7 ( see corollary [ coxdiag ] ) .
we assume that diagram nodes are numbered 1,2,3 left to right , that each node represents the corresponding reflection @xmath65 , @xmath66 , @xmath67 acting in @xmath11 .
decorations described in this subsection should be viewed superimposed on any of the four diagrams ( cor .
[ coxdiag ] ) .
let the seed point ( face @xmath34 ) be either @xmath68 or @xmath69 , corresponding to the dual pair of polytopes .
we denote by @xmath27 the reflections that move the seed point and by @xmath28 the reflections that stabilize it .
consider two seed point decorations of all four diagrams ( cor .
[ coxdiag ] ) , namely @xmath70 here , @xmath27 stands for the seed point reflection .
the sub - diagram decorated by @xmath71 identifies reflections generating @xmath72 .
the symmetry group @xmath73 of the face @xmath34 is trivial , the face is just a point . the orders of reflection groups @xmath74 in @xmath11 are shown in table [ rootnumbers ] .
the groups @xmath75 stabilizing @xmath68 are respectively latexmath:[\[\begin{gathered } based on the number of vertices , we have a tetrahedron for @xmath12 , an octahedron for @xmath13 and @xmath14 , and a dodecahedron for @xmath7 . the groups @xmath75 stabilizing @xmath69 are all of type @xmath77
hence , the numbers of vertices are those of a tetrahedron for @xmath12 , a cube for @xmath13 and @xmath14 , and a icosahedron for @xmath7 .
the dual tetrahedra of @xmath12 are differently oriented in @xmath11 , but otherwise are identical .
the second step describes the edges of the solids .
the decorations are @xmath78 the symmetry group generated by @xmath65 refers to the edge with end points @xmath68 and @xmath79 . for the dual polytope ,
the symmetry group is generated by @xmath67 , and it refers to the edge with end points @xmath69 and @xmath80 .
note that this description of edges is independent of the coxeter groups used in cor .
[ coxdiag ] . in all cases ,
the group @xmath81 is @xmath82 of order 4 .
consequently , the number of edges is @xmath83 .
we get 6 for @xmath12 ( tetrahedron ) , 12 for @xmath13 and @xmath14 ( cube and octahedron ) and 30 for @xmath7 ( icosahedron and dodecahedron ) .
the final step in the decoration @xmath84 describes the 2-dimensional faces @xmath85 .
the sub - diagram @xmath86 in is the group @xmath87 of order @xmath88 for all cases . @xmath89 the same sub - diagram in stands for @xmath87 in the case of @xmath12 , @xmath90 for @xmath13 and @xmath14 , and @xmath91 for @xmath7 .
therefore @xmath92 the shape of face @xmath85 can be easily determined from the relative angles of the mirrors @xmath65 and @xmath66 acting on @xmath68 or from reflections @xmath66 and @xmath67 acting on @xmath69 .
a summary of properties of platonic solids in @xmath11 and their faces is shown in table [ platodim3 ] .
@xmath93 the dual pair of platonic solids of @xmath12 consists of two tetrahedra differently oriented in @xmath11 , their seed points being @xmath68 and @xmath69 .
the vertices of the two tetrahedra are the following @xmath94 connected coxeter - dynkin diagrams of the groups generated by four reflections are of types @xmath16 , @xmath17 , @xmath18 , @xmath6 and @xmath8 ( see ( corollary [ coxdiag ] ) ) . using table [ platodim3 ]
conventions , properties of the 4-dimensional platonic solids , together with their duals , are summarized in table [ platodim4 ] .
@xmath95 [ edgmeetver ] let us answer the question posed in the introduction : how many edges @xmath96 meet in a vertex @xmath34 of any platonic solid ?
decorations of coxeter - dynkin diagrams for 4-dimensional platonic solids and their duals are : @xmath97 from table [ platodim4 ] , we read the stabilizers of faces @xmath98 the edges originating in @xmath34 or @xmath99 are generated by the stabilizer of @xmath34 or @xmath99 , respectively .
they are equal to @xmath100 for , and @xmath101 for .
the formula for the number of edges meeting in one vertex can be written as @xmath102 or more generally , if we consider @xmath103 of faces @xmath10 having in common faces @xmath104 for @xmath105 : [ numvert ] the number of faces @xmath10 meeting in a face @xmath104 for @xmath106 is equal to the ratio of orders of coxeter groups of the subgroups @xmath107 which stabilize a given face @xmath108 , respectively @xmath109 the groups @xmath110 for are respectively @xmath111 and @xmath112 , and for @xmath113 .
table [ edges4 ] summarizes this proposition .
@xmath114 both the pentatope and 24-cell are self - dual , the 16-cell is dual to the tessaract , and the 600-cell and 120-cell are dual to each other @xcite .
the are only 3 types of platonic solids for dimension more than 4 , namely the simplex , hypercube and cross - polytope .
the simplex is self - dual , and the cross - polytope and hypercube are dual to each other . the corresponding coxeter - dynkin diagrams ( cor . [ coxdiag ] ) are of types @xmath116 and @xmath5 .
using table [ platodim3 ] conventions , the properties of platonic solids of dimension @xmath19 , together with their duals , are summarized in table [ platodimn ] .
@xmath117 table [ meetingfaces ] summarizes this proposition [ numvert ] for any n - dimensional polytope .
@xmath118 analog to proposition [ numvert ] , we can give the formula for finding the number of faces @xmath10 meeting at face @xmath119 , where @xmath120 .
the number of faces @xmath10 meeting at face @xmath121 for @xmath122 is equal the ratio of orders of weyl groups of subgroups @xmath123 , which stabilize a given face @xmath119 and @xmath124 , respectively @xmath125 [ facemeet ] we now consider the 4-dimensional faces appearing in 5-dimensional polytopes .
diagrams of the face @xmath126 are : @xmath127 @xmath128 the shape of face @xmath126 is determined from the relative angles of mirrors @xmath129 acting on @xmath68 , and from reflections @xmath130 acting on @xmath131 . using tables [ platodim4 ] and [ edges4 ] , one can see from sub - diagram @xmath132 that , for @xmath133 , faces are pentatopes , and for @xmath134 , either 16-cells or tessaracts .
a seed decoration is often applied to several coxeter diagrams .
see an example in [ solid4dim ] . the same is true for recursive decorations that follow from the seed .
therefore , the set of such decorations could be viewed as a ` generic polytope ' pertinent to all coxeter groups with the same connectivity of their coxeter - dynkin diagrams .
can anything be learned from the generic set of decorations ? in this paper , we focused on coxeter groups whose diagrams are connected . in general , the decoration rules can be used for polytopes of lie algebras that are semisimple but not simple , i.e. their coxeter - dynkin diagrams are disconnected . in this case , each connected component must have its own seed point , indicated in the initial decoration .
consequently , there are several orbits of edges , hence , such polytopes are never platonic . in recent years evidence
has been obtained that there exist in nature molecules with imperfect symmetries @xcite of @xmath22 type , not necessarily platonic .
they could be considered as @xmath22 symmetries broken to a subgroup .
r. v. moody , j. patera , _ voronoi domains and dual cells in the generalized kaleidoscope with applications to root and weight lattices , _
( dedicated to h. s. m. coxeter ) , can .
j. math .
, * 47 * ( 1995 ) 573605 r. v. moody , j. patera , _ voronoi and delaunay cells of root lattices : classification of their faces and facets by coxeter - dynkin diagrams , _
( dedicated to the memory of h. zassenhaus ) , j. phys .
a : math . gen . , * 25 * ( 1992 ) 5089 - 5134 r. v. moody , j. patera , _ a coxeter diagram description of faces of voronoi cells and their duals , _ proc .
group theor . meth . in phys . , salamanca , 1992 , group theoretical methods in physics , vol . 1 , p.438 - 443 , anales de fsica , ( 1993 ) eds
. m. del olmo , m. santander , j. guilarte | this paper considers platonic solids / polytopes in the real euclidean space @xmath0 of dimension @xmath1 .
the platonic solids / polytopes are described together with their faces of dimensions @xmath2 .
dual pairs of platonic polytopes are considered in parallel .
the underlying finite coxeter groups are those of simple lie algebras of types @xmath3 , @xmath4 , @xmath5 , @xmath6 and of non - crystallographic coxeter groups @xmath7 , @xmath8 .
our method consists in recursively decorating the appropriate coxeter - dynkin diagram .
each recursion step provides the essential information about faces of a specific dimension . if , at each recursion step , all of the faces are in the same coxeter group orbit , i.e. are identical , the solid is called platonic .
centre de recherches mathmatiques , universit de montral , c. p. 6128 centre ville , montral , h3c3j7 , qubec , canada .
permanent address : institute of mathematics , university of bialystok , akademicka 2 , pl-15 - 267 , bialystok , poland .
e - mail : [email protected] keywords : platonic solids , coxeter group , coxeter - dynkin diagram , lie groups/ algebras msc : 97g40 , 20e32 , 22d05 , 17b22 |
superprocesses in random environments were introduced in @xcite as the scaling limits of particle systems whose branching are affected by random environments . in particular the limiting behavior of the following model has been studied . at time @xmath1 ,
@xmath2 particles are located in @xmath3 .
each of these @xmath4 particles follows the path of an independent brownian motion until time @xmath5 . at time @xmath6
each particle independently of the others either splits into two or dies and then the individual particles in the new population again follow the paths of independent brownian motions starting at their place of birth , in the interval @xmath7 , and the pattern of alternating branching and spatial spreading continues .
let us describe in details the branching mechanism that was suggested in @xcite .
let @xmath8 be a sequence of i.i.d .
@xmath3-indexed random fields with mean @xmath9 and covariance @xmath10 at time @xmath11 each particle , independently of the others conditionally on @xmath12 , either splits into two with probability @xmath13 or dies with probability @xmath14 where @xmath15 is the location of the particle .
that is , the fields @xmath16 create the random environment that affects the branching of the particles .
define the following measure - valued process that describes the evolution of the population : @xmath17 before proceeding we introduce some notation . for a locally compact polish space @xmath18 ,
let @xmath19 ( respectively , @xmath20 ) be the space of finite ( respectively radon ) non - negative measures on @xmath18 , equipped with the weak ( respectively , vague ) topology ( see section 3.1 in @xcite ) . in the case of @xmath21 ,
we will also write @xmath22 and @xmath23 . both @xmath24 and @xmath25 denote the integral of a function @xmath26 with respect to measure @xmath27 . for any metric space @xmath18 let @xmath28 ( resp .
@xmath29 ) be the space of cadlag ( resp .
continuous ) @xmath18-valued functions on @xmath30 endowed with the skorohod topology .
let @xmath31 ( resp .
@xmath32 ) be the set of continuous ( resp . bounded continous ) functions with continuous ( resp . bounded continuous ) partial derivatives of order @xmath33 or less . also we define @xmath34 to be the set of bounded measurable functions on @xmath3 .
it was shown in @xcite , under some additional technical assumptions on @xmath12 , that if @xmath35 then @xmath36 here @xmath37 is a process in @xmath38 which is the unique solution to the following martingale problem : @xmath39 , @xmath40 is a continuous martingale with quadratic variation process @xmath41 in this paper we introduce some minor changes into the above model . instead of the binary branching
we assume that each particle gives birth to a number of particles distributed according to the geometric distribution with parameter @xmath42 ; that is , if @xmath43 is the number of offspring of the particle located at @xmath15 at time @xmath44 , then @xmath45 in particular , conditioned on the environment @xmath12 , the expected number of offspring of a particle at @xmath15 at time @xmath44 is @xmath46 compared with @xcite , we also allow @xmath12 to be slightly more general , that is , we assume that @xmath47 is a sequence of i.i.d .
random fields with mean @xmath48 , for some @xmath49 , and covariance @xmath50 let @xmath51 be defined for this model as in ( [ eq:1.1.1 ] ) . by the same argument as in
@xcite one can prove that the limit of @xmath52 is the ( unique ) solution to the following martingale problem : @xmath53 , @xmath54 is a continuous martingale with quadratic variation process @xmath55 where @xmath56 .
( note the different factor multiplying the term @xmath57 here compared with , which comes from the different variances of the geometric and bernoulli distributions . )
the main purpose of this paper is to study the brownian snake representation of the process that solves the above martingale problem ( [ mp-1]-[mp-2 ] ) .
for a nice introduction into the topic the reader is referred to @xcite . the classical brownian
snake was used to study different properties of super - brownian motion . loosely speaking if @xmath58 is a brownian snake then for each @xmath59
, @xmath60 is a stopped brownian path .
to be more precise we call the pair @xmath61 a stopped path in @xmath3 if for each @xmath62 , @xmath63 .
@xmath64 is called the lifetime of the path @xmath65 and sometimes is denoted by @xmath66 or @xmath67 .
let @xmath68 denote the space of all stopped paths in @xmath3 equipped with the distance @xmath69 we will also use the notation @xmath70 for the terminal point of @xmath65 . for any @xmath71 we denote by @xmath72 the path with lifetime @xmath9 constantly equal to @xmath15 .
if @xmath73 is a stopped path then with some abuse of notation we will sometimes set @xmath74 for any @xmath59 . the usual brownian snake can be thought of as a limit of the so - called discrete snakes that we will now define .
let @xmath75 be a rescaled simple random walk on @xmath76 reflected at the origin , that is , the time between the steps is @xmath77 and the size of the jump is @xmath78 with equal probabilities .
explicitely , @xmath79 we also let @xmath80 be constant between the jumps .
the process @xmath80 is called the contour or lifetime process of the discrete snake @xmath81 .
that is for each @xmath59 , the snake @xmath82 is a stopped path with life time @xmath83 .
we next define the paths of the snake .
fix @xmath71 and set @xmath84 let @xmath85 be a sequence of independent brownian paths stopped at time @xmath6 , independent of the contour process .
let @xmath86 be the stopped path at time @xmath87 with lifetime @xmath88 . then define @xmath89 where @xmath90 denotes the concatenation of two paths @xmath91 and @xmath92 in the obvious way . in words
, if the lifetime @xmath93 goes down by @xmath6 we erase the path of the snake from the tip by @xmath6 , or to put it differently , we reduce its lifetime by @xmath6 . if @xmath93 goes up by @xmath6 we add the path @xmath94 to the tip of the snake .
then we define @xmath95 this way we constructed a sequence of discrete snakes . as is the case for @xmath80
, we define @xmath96 .
the sequence of processes @xmath81 converges , as @xmath97 , to a continuous time brownian snake ( see e.g. proposition 2.2 in @xcite ) .
we next describe the connection between the snake process and the branching brownian motion .
define the discrete version of the local time as the rescaled number of upcrossings of @xmath93 from the corresponding level : @xmath98 we also define , for @xmath99 , @xmath100 since @xmath101 is increasing we define the measure @xmath102 in an obvious way .
in fact this convention will be used throughout the paper : for any non - decreasing function @xmath103 on @xmath104 , @xmath105 , with a slight abuse of notation , will denote the corresponding measure defined via @xmath106)=f_b - f_a$ ] , for any @xmath107 . for
any @xmath108 introduce the inverse local time at level @xmath109 as @xmath110 for any @xmath111 and @xmath112 define the measure valued process @xmath113 so that , for @xmath114 @xmath115 it is easy to see that @xmath116 , is the measure - valued process constructed in the previous section starting at `` time '' @xmath117 such that @xmath118 and therefore @xmath119 where @xmath120 solves the martingale problem starting at time @xmath109 such that @xmath121 and , @xmath53 , @xmath122 is a continuous martingale with quadratic variation process @xmath123 we finally define the discrete snake in random environment corresponding to the branching processes in random environment described in section [ 1.2 ] .
the main difference with the `` fixed environment '' case is that here the snake can not be constructed conditionally on the lifetime process .
both processes have to be constructed simultaneously . the environment @xmath124 is assumed to consist of a sequence of i.i.d .
random fields , satisfying @xmath125 and @xmath126 , with mean @xmath48 , for some @xmath49 , and covariance @xmath127 as in ( [ eq - nov26a ] ) , with @xmath128 .
now define the snake with lifetime processes @xmath129 as follows .
fix a constant @xmath130 .
let @xmath131 and @xmath132 with @xmath133 .
suppose we are given @xmath134 for some @xmath135 .
@xmath136 will be defined as follows . if @xmath137 , then conditionally on @xmath12 and @xmath138 we set @xmath139 where we introduced above the notation for the `` tip '' of the snake : @xmath140 if @xmath141 , then with probability one we set @xmath142 .
if @xmath143 , then with probability one we set @xmath144 . ( that is , the process is reflected at height @xmath145 ; a similar approach of introducing a super - critical branching mechanism via a reflection of the lifetime process was used by j .- f .
delmas in @xcite . )
let @xmath85 be a sequence of independent brownian motions stopped at time @xmath6 .
given the evolution of the lifetime process @xmath93 until time @xmath146 , the path of the brownian snake @xmath147 at time @xmath146 is defined exactly as in ( [ eq:1.8 ] ) .
we next explain the connection between the snake and branching particle system in random environment which is analogous to the connection that exists between the processes in a constant environment .
define the rescaled local time @xmath148 for @xmath93 as in ( [ eq:1.3 ] ) , ( [ eq:30_12_10 ] ) and the inverse local time as in ( [ eq:1.4 ] ) . for any @xmath149 , we define the measure - valued process in the same way as it is done in ( [ eqm40 ] ) : @xmath150 for all @xmath114 .
this process characterizes the branching particle picture in random environment with offspring distribution given by ( [ eq:1.2 ] ) and starting with @xmath151 particles at the site @xmath71 at time @xmath152 . in the case of @xmath153
, we will use the notation @xmath154 for the corresponding process .
the following is our first main result .
[ thm:1 ] fix @xmath130
. then the sequence of processes @xmath155 @xmath156 @xmath157 is @xmath158-tight in @xmath159 .
let @xmath160 be an arbitrary limiting point , let @xmath161 be a local time of @xmath162 at level @xmath109 and let @xmath163 be the inverse of the local time . fix an arbitrary @xmath164
. then @xmath165,\end{aligned}\ ] ] is the measure - valued process satisfying the martingale problem ( [ mp-1]-[mp-2 ] ) on @xmath166 $ ] , with @xmath167 .
( the uniqueness of solutions to can be read off @xcite , however the uniqueness of the snake @xmath147 is an open problem . ) in the particular case of a spatially smooth " random environment we can give another description of the snake process .
it is easy to check from our assumptions on @xmath168 that if we define @xmath169 then @xmath170 where @xmath171 is a gaussian generalized noise on @xmath172 , white in time and colored in space , such that @xmath173 where @xmath174 is the dirac measure at @xmath9 . given the result on the tightness of @xmath155 , one can easily deduce that the pair @xmath175 is tight . in what follows
we assume that @xmath176 is a limit point of the tight sequence @xmath175 , and we recall that @xmath160 . our aim is to introduce a particular functional of the limiting snake that has a simple semimartingale decomposition .
the definition of the functional is motivated by the one used by dhersin and serlet @xcite and also by a functional used to transform brox s diffusion into a martingale , see @xcite . for @xmath177 , let @xmath178 our second main result is the following . for technical reasons , we restrict attention to branching laws constructed directly from smooth gaussian fields .
[ thr:5.1 ] fix @xmath130 .
let @xmath179 be a gaussian field as in so that @xmath180 , a.s .. , and set @xmath181 let @xmath176 denote a limit point of @xmath182 .
then @xmath183 has the distribution of @xmath179 , and there exists a brownian motion @xmath184 such that @xmath185 the first term on the right side of ( [ eq:5.1 ] ) can be written as @xmath186 and it comes from the fact that @xmath187 is a brownian path .
it is plausible that one may relax the assumptions on the fields @xmath188 in theorem [ thr:5.1 ] ; however , some strong assumptions that ensure good spatial approximation of @xmath183 by partial sums of the @xmath168s seem to be crucial .
note that in the case of constant function @xmath189 , for every @xmath190 , @xmath191 is a constant function in space , and hence we immediately have the following corollary , which for simplicity we state only in case @xmath192 . a similar result ( without the reflection )
can be found in @xcite .
[ cor - brox ] let @xmath193 and @xmath192 .
then @xmath162 is the brox diffusion reflected at @xmath9 and @xmath145 .
see the appendix for the definition of the brox diffusion . in the next section ,
we derive some standard estimates on survival probability for branching processes in a random environment .
section [ sec - tcont ] is concerned with the proof of tightness of the contour process .
( because of dependence through the environment , natural arguments involving stopping times such as aldous tightness criterion can not be applied directly , and extra care has to be employed in separating dependence on the level of the contour process from dependence on the lifetime of the process . )
most of the work is devoted to proving that large upward jumps of the contour process are unlikely ; downward jumps are then handled by a time reversal argument .
section [ sec-4 ] is devoted to the proof of tightness of the snake process and its local time , and a completion of the proof of theorem [ thm:1 ] .
section [ sec-5 ] is devoted to the description of the snake provided in theorem [ thr:5.1 ] , while the appendix is devoted to the description of the contour process for environments with no spatial dependence , providing in particular a direct proof of corollary [ cor - brox ] , that bypasses the need to consider the brownian snake .
* notation * throughout , @xmath194 denote generic constants whose values may change from line to line .
numbered constants ( such as @xmath195 , etc . ) are fixed and do not change throughout the paper .
we start with a lemma that describes the asymptotics for survival probability for classical branching processes . for any @xmath196 let @xmath197 be the branching process with geometric offspring distribution with parameter @xmath198 for some @xmath199 .
that is if @xmath200 is the number of offspring in the process @xmath201 , then @xmath202 for @xmath203 define @xmath204 [ lem:3 ] assume @xmath205 and @xmath206 for all @xmath196 . then for any @xmath203 , @xmath207}>0)= h(b,\delta).\ ] ]
[ [ proof ] ] proof : + + + + + + for @xmath208 the result is well - known ( see e.g. ( * ? ? ? * theorem ii.1.1 ) for a more general result ) .
while we believe that the result is also known for @xmath209 , we were unable to locate a reference and thus for the sake of completeness we provide a proof .
let @xmath210 be the generating function of @xmath200 , that is @xmath211 define @xmath212 and in general @xmath213 then by the branching property , @xmath214 = f_k(s).\ ] ] therefore , @xmath215 fix @xmath216 and define @xmath217 one has @xmath218 let @xmath219 then @xmath220 and @xmath221 with @xmath222 by iterating we get @xmath223 now as @xmath224 we have @xmath225 also recall that @xmath216 and hence @xmath226 therefore @xmath227 since @xmath228}>0)\sim \frac{y_{\lfloor n\delta\rfloor}}{\delta}= \frac{1}{\delta z_{\lfloor n\delta\rfloor}}$ ] , this concludes the proof .
width5pt returning to the random environment case , let @xmath229 denote the total mass of the branching brownian motion in random environment @xmath51 ( with geometric offspring distribution ) defined in section [ 1.1 ] ( that is , @xmath230 ) and let @xmath201 be as above with @xmath231 [ rem - new ] with our assumptions , it is easy to see that the @xmath232-th moment ( for any @xmath233 ) of the absolute value of the expected ( conditioned in the environment ) number of offspring minus 1 of a particle at @xmath15 at time @xmath44 , see ( [ eqm32 ] ) , is bounded by @xmath234 , for an appropriate constant @xmath235 .
moreover , the absolute value of the first moment of the number of offspring minus 1 of a particle at @xmath15 at time @xmath236 , see ( [ eqm32 ] ) , is bounded by @xmath237 , for an appropriate constant @xmath238 .
[ lem:4 ] let @xmath239 for all @xmath196 .
then @xmath240 [ [ proof-1 ] ] proof : + + + + + + for @xmath241 we denote by @xmath242,$ ] the position at time @xmath243 of the @xmath244-th particle that was born at time @xmath44 .
that is we have @xmath245 moreover if @xmath246 is the number of offspring at time @xmath247 of the @xmath244-th particle that was born at time @xmath44 , then we also have @xmath248 we write for simplicity @xmath249 and denote by @xmath250 the sigma - algebra generated by the environment @xmath251 .
we have , for @xmath252 , @xmath253 therefore , @xmath254}\right ) \right ) , \nonumber\end{aligned}\ ] ] where the last inequality follows by jensen inequality .
since @xmath255 we get by trivial estimates that for @xmath256 large enough @xmath257 again by trivial estimate on the logarithmic function we get @xmath258 for all @xmath256 sufficiently large .
take an expectation to get @xmath259 substituting in ( [ equt:03_2 ] ) we get @xmath260 let @xmath210 be the generating function of the geometric distribution with parameter @xmath261 , then @xmath262 if one takes @xmath263 then we get that @xmath264 and hence by iterating ( [ equt:03_5 ] ) we get @xmath265 therefore @xmath266 and hence by lemma [ lem:3 ] we get that @xmath267 height5pt width5pt [ lem:5 ] let @xmath268 be as above .
* for any @xmath203 , @xmath269 and hence , @xmath270 * for any @xmath271 , @xmath272 [ [ proof-2 ] ] proof : + + + + + + * ( a ) * the proof of goes along the similar lines as the proof of the previous lemma .
first recall that @xmath273 for all @xmath274 .
hence , by iteration , we get @xmath275 and the result follows .
+ * ( b ) * for all @xmath135 , define @xmath276 then using ( [ eq:28_12_1 ] ) it is easy to check that @xmath277 is a nonnegative @xmath278-supermartingale .
therefore by maximal inequalities for non - negative supermartingales we get @xmath279 to prove the result we consider the cases @xmath280 and @xmath281 separately .
first suppose that @xmath282 .
recall the definition of @xmath283 to get that , in this case , @xmath284 by putting ( [ eq:28_12_2 ] ) , ( [ eq:28_12_3 ] ) together we get that @xmath285 now let @xmath286
. then we get @xmath287 apply this and ( [ eq:28_12_1 ] ) with @xmath288 instead of @xmath109 to get @xmath289 by combining ( [ eq:28_12_4 ] ) , ( [ eq:28_12_6 ] ) , we get @xmath290 and by letting @xmath224 we are done .
width5pt the next result generalizes the previous lemma .
[ lem : nov1 ] let @xmath291 be a bounded non - negative measurable function on @xmath3 .
then , for any @xmath203 , @xmath292 where @xmath293 is the semigroup of the brownian motion .
[ [ proof-3 ] ] proof : + + + + + + the proof goes along the similar lines as the proof of the previous lemma . for any @xmath294 we have @xmath295 for all @xmath296 and
hence by iteration @xmath297 and the result follows .
in this section we will prove the tightness of the sequence of the contour processes @xmath298 .
the following proposition is the main result of this section .
[ prop : tight ] for any @xmath203 , @xmath299 , @xmath300 that is , @xmath298 is @xmath158-tight in @xmath301 .
the proof of the proposition will be given in this section .
recall the definition of the discrete version of the local time for @xmath93 and its inverse ( see ( [ eq:1.3 ] ) , ( [ eq:30_12_10 ] ) and ( [ eq:1.4 ] ) for the same definitions in the case without environment ) .
fix an arbitrary @xmath302 .
we will first handle the tightness on the time interval @xmath303,\ ] ] and we start with the following proposition .
[ prop:1 ] for any @xmath203 , @xmath304 the proof of the proposition will be given after we present several preliminary lemmas .
for any @xmath108 and @xmath112 , recall the measure - valued process @xmath305 , see ( [ eq:31_12_01 ] ) . fix an arbitrary @xmath203 . recall that @xmath306 was defined in ( [ equt : nov6 ] ) .
we have the following lemma . [ lem3b2 ] for any @xmath307 @xmath308 [ [ proof-4 ] ]
proof : + + + + + + @xmath309 since , by lemma [ lem:4 ] , @xmath310 the result follows .
width5pt the following corollary is immediate .
[ cor:2 ] for any @xmath307 @xmath311 the next corollary gives a bound on the positive increment of @xmath93 .
[ cor:3 ] for any @xmath312 @xmath313 [ [ proof-5 ] ] proof : + + + + + + we first prove that for any @xmath314 , @xmath315 suppose that there exist deterministic @xmath316 and subsequences @xmath317 such that @xmath318 to avoid cumbersome notation , for the rest of the proof we write @xmath256 and @xmath319 for @xmath320 and @xmath321 respectively .
note that it follows from ( [ equt:2 ] ) that @xmath322 then as in lemma [ lem3b2 ] , we may define the sequence of measure - valued processes @xmath323 with total mass @xmath324 this process starts at the total mass @xmath325 and , appealing to @xcite , as @xmath224 , it converges weakly in @xmath301 to the continuous process @xmath326 starting at @xmath327 .
therefore , by the weak convergence properties , there exists @xmath328 such that @xmath329 note that on the event in ( [ eq3b13 ] ) , we have @xmath330 summing over @xmath331 , we get that with probability greater than @xmath332 , the occupation time of @xmath333 on the time interval @xmath334 $ ] is bounded from below by @xmath335 on the other hand the total occupation time of @xmath333 on the interval @xmath336\subset[\tau^{n , a}_r,\tau^{n , a}_r+\ep_n]\ ] ] is bounded by @xmath337 , which contradicts ( [ equt:1 ] ) .
hence ( [ equt:2 ] ) follows .
continuing with the proof of the lemma , we have from ( [ equt:2 ] ) that for any @xmath314 , @xmath338 where the last inequality follows by ( [ equt:2 ] ) and corollary [ cor:2 ] . since @xmath327 was arbitrary we are done . ' '' '' we now introduce further notation .
let @xmath339 we will prove the following lemma . for any @xmath203 , [ lem:2 ]
@xmath340 [ [ proof-6 ] ] proof : + + + + + + now we will need some further notation .
denote @xmath341$}\}\\ & = & \{\mbox{number of particles in the original branching particle system } \\ & & \mbox{at time $ a_nn$ whose descendants survive till time $ ( a_n+s)n$}\}\end{aligned}\ ] ] by lemmas [ lem:4 ] , [ lem:5 ] and the markov property of the branching system we immediately get @xmath342 for @xmath343 define @xmath344 where @xmath345 that is , @xmath346 are the times when successful excursions of @xmath93 reach the level @xmath347
. then we have , for any fixed integer @xmath232 , @xmath348 by an argument similar to the one in corollary [ cor:3 ] we get that @xmath349 this implies that @xmath350 where the last inequality follows by the markov inequality and ( [ equt:03_6 ] ) .
since @xmath232 was arbitrary we are done . ' '' '' we can now complete the proof of proposition [ prop:1 ] .
[ [ proof - of - propositionprop1 ] ] proof of proposition [ prop:1 ] : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + for @xmath203 let @xmath351\}.\end{aligned}\ ] ] then @xmath352 however by lemma [ lem:2 ] we get that , for every @xmath244 , @xmath353 and this finishes the proof of proposition [ prop:1 ] . ' '' '' to handle downward jumps , we need the following proposition .
[ prop:2 ] for any @xmath203 , @xmath354 [ [ proof-7 ] ] proof : + + + + + + in fact the proof is easy if one considers the process @xmath93 reversed in time , that is the process @xmath355 , which is easily seen ( see the explicit argument below ) to possess the same law ( with @xmath356 ) as the original process @xmath80 . since any jump down for @xmath93 becomes a jump up for @xmath357 , the claim ( [ of-220209 ] ) follows from proposition [ prop:1 ] applied to @xmath357 . to see the reversibility claim ,
we introduce a sequence of path transformations @xmath358 on @xmath359 , each of which is measure preserving and preserves @xmath360 , such that @xmath361 where @xmath362 denotes the image of @xmath363 under the transformations .
this will prove the claim .
to avoid cumbersome notation , we consider the case of @xmath364 only , and we omit the index @xmath256 .
the general @xmath256 can be treated the same way with proper scaling . for @xmath365 , the transformation @xmath366
is obtained as follows .
if @xmath367 for some integer @xmath368 , that is @xmath243 is a return time of @xmath369 to @xmath9 , then define @xmath370 .
if @xmath371 for some @xmath372 , that is @xmath243 belongs to the @xmath373th excursion of @xmath369 from @xmath9 , then define @xmath374 .
then , @xmath375 ( in words , @xmath366 reverses the order of the excursions from @xmath9 but keeps the time orientation of each excursion intact ; thus , the total length of the excursions is preserved . )
it is straightforward to check that the law of @xmath376 is the same as that of @xmath377 . for @xmath378 ,
let @xmath379 and for @xmath380 , @xmath381 @xmath382 is then defined as @xmath366 applied to the excursions of the path _ from level @xmath383_. explicitly , let @xmath384
. for @xmath385\right)^c$ ] , set @xmath386 .
for each @xmath373 , let @xmath387 , @xmath388 , and @xmath389 .
let @xmath390 be defined on the interval @xmath391 in the same way as the case of @xmath365 with @xmath392 replaced by @xmath393 . then , @xmath394 again , in words , @xmath395 reverses the order of the excursions from @xmath383 but keeps the time orientation of each excursion intact ; thus , the total length of the excursions is preserved . )
it is straightforward to check that the law of @xmath396 is the same as that of @xmath377 .
we can continue this procedure for @xmath397 . as explained above
, this completes the proof . ' '' '' to finish the proof of the proposition [ prop : tight ] we need the folowing lemmas that describe the limiting behavior of @xmath398 and @xmath399 ( recall that @xmath400 was introduced in ( [ 01_06 ] ) ) .
[ lem:01_06_1 ] for any @xmath302 , the sequence of processes @xmath401 is @xmath158-tight in @xmath402 . [ [ proof-8 ] ] proof : + + + + + + first recall from ( [ 01_06 ] ) and ( [ eq : oc5_1 ] ) , that @xmath403 is the total mass at time @xmath109 of the measure - valued process @xmath404 defined in the introduction . since the sequence of measure - valued processes @xmath405 is @xmath158-tight in @xmath406 ( see @xcite and the comments leading to ( [ mp-3 ] ) ) , we get the desired result . ' '' '' the next lemma studies the limiting behavior of @xmath399 . toward this end ,
recall that according to our conventions introduced after ( [ eq:30_12_10 ] ) , we use the same notation for an increasing function and the corresponding measure .
[ lem:01_06_2 ] * for any @xmath164 , the sequence of random variables @xmath407 is tight and any limit point @xmath408 satisfies @xmath409 * for any @xmath410 , @xmath411 , there exists @xmath412 , such that @xmath413 * the sequence @xmath414 is tight in @xmath415 .
* let @xmath416 be an arbitrary limit point of @xmath399 .
then for any fixed @xmath417 , @xmath418 is continuous at @xmath419 with probability @xmath420 .
[ [ proof-9 ] ] proof : + + + + + + * ( a ) * define @xmath421 note that @xmath422 on the other hand @xmath423 and since by lemma [ lem:01_06_1 ] , @xmath424 is tight , by ( [ 01_06_2 ] ) we get the tightness of @xmath425 .
similarly , since @xmath424 is @xmath158-tight for any @xmath410 we can fix @xmath426 such that @xmath427 for all @xmath256 sufficiently large . using this , ( [ 01_06_2 ] ) and the definition of @xmath428 we get @xmath429 with probability at least @xmath430 for all @xmath256 sufficiently large .
since @xmath431 was arbitrary we get that any limit point of @xmath432 is greater than @xmath9 with probability @xmath420 . + * ( b ) * for any @xmath433 we can represent @xmath434 where , for each @xmath244 , @xmath435 is distributed as @xmath432 . fix arbitrary @xmath436 . since , by part * ( a ) * of the lemma , any limit point of @xmath435 is strictly greater than @xmath9 with probability one , we can easily choose @xmath437 sufficiently large such that @xmath438 with probability at least @xmath439 , for all @xmath256 sufficiently large .
+ * ( c ) * immediate from * ( a)*. + * ( d ) * let @xmath416 be a limiting point @xmath399 . to prove this part of the lemma we have to show that , for any @xmath410 , there exists @xmath203 , such that @xmath440 similarly to what we have done in * ( a ) * define , @xmath441 then we have @xmath442 where recall that @xmath443 is the measure - valued process corresponding to the branching particle system in random environment , constructed in section [ sec:1 ] ( see ( [ eq:31_12_01 ] ) ) , that starts at time @xmath444 with initial mass @xmath445 . by lemma [ lem:5](b ) @xmath446 for all
@xmath256 sufficiently large .
we can take @xmath426 sufficiently small such that the right hand side of ( [ eq:28_12_8 ] ) is less than @xmath447 , and this together with ( [ eq:28_12_9 ] ) implies that @xmath448 for all @xmath256 sufficiently large .
therefore ( [ eq:28_12_10 ] ) follows for any limit point of @xmath399 . ' '' ''
now we are ready to complete the proof of proposition [ prop : tight ] .
[ [ proof - of - propositionproptight ] ] proof of proposition [ prop : tight ] : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + proposition [ prop : tight ] follows immediately from propositions [ prop:1 ] , [ prop:2 ] , lemma [ lem:01_06_2]*(b ) * , and the fact that @xmath449 was arbitrary . ' '' ''
the bulk of this section is devoted to the proof of the following proposition .
[ prop:3 ] the sequence @xmath451 is tight in @xmath452 let @xmath453 be its arbitrary limiting point
. then @xmath453 belongs to @xmath454 .
moreover , @xmath455 is the local time of @xmath162 ( @xmath162 is the lifetime of @xmath147 ) , that is , @xmath456 note that following our conventions , we denote by @xmath457 the measure and by @xmath458)$ ] the corresponding increasing distribution function corresponding to @xmath455 .
the proof of proposition [ prop:3 ] is long and we indicate the main steps
. we will first prove the tightness of the sequence of processes @xmath155 , based on the tightness of the contour process established in section [ sec - tcont ] .
this will be obtained in lemma [ lem:07_06_1 ] , after going through a fair amount of preliminary material .
the tightness of the sequence of the local time process @xmath459 is then obtained in lemma [ lem - newtl ] , thus completing the proof of proposition [ prop:3 ] .
the rest of the section is devoted to the identification of the limiting snake representation .
here we have to identify a limit point of the sequence of the local times @xmath459 as the local time of the limiting contour process , and this is done in lemma [ lem - loctime ] .
additionally , in lemma [ lem:28_12_1 ] we verify that a limiting point of @xmath460 is indeed the value at @xmath449 of the inverse function of the limiting local time .
the proof of theorem [ thm:1 ] is an immediate corollary of these facts , and is presented at the end of the section . as in the previous section , where the tightness of the contour processes @xmath298
was obtained , we first handle tightness on the time interval @xmath461 $ ] .
fix an arbitrary @xmath462 and recall that @xmath463 ( see ( [ eq : oc5_1 ] ) ) is the measure - valued process characterising the branching particle picture , and in particular , @xmath464 is the number of particles alive at time @xmath465 .
first we derive a bound on the maximal displacement of the offsprings from the ancestors during the time interval @xmath466 $ ] .
this estimate will be crucial for proving tightness of paths of the brownian snake in random environment .
fix @xmath467 arbitrary small .
define @xmath468 [ lem : nov4 ] there exists @xmath469 , such that @xmath470 we postpone the proof of lemma [ lem : nov4 ] , and prepare some preliminary estimates .
introduce the event @xmath471 and define @xmath472 which gives the number of particles alive at time @xmath473 whose historical paths were displaced by distance more than @xmath474 on the time interval @xmath475 $ ] .
[ lem : nov2 ] there exist @xmath476 and @xmath477 such that , for all @xmath256 sufficiently large , @xmath478 [ [ proof-10 ] ] proof : + + + + + + let @xmath479 that is , @xmath480 is the total number of particles that are alive at time @xmath481 and whose historical paths were displaced by distance more than @xmath474 on the time interval @xmath475 $ ] .
we enumerate these particles by @xmath482 and let @xmath483 be the number of living descendents of the particle @xmath244 ( @xmath484 ) at time @xmath473 . then clearly @xmath485 lemma [ lem:4 ] and ( [ eq - nov26c ] ) imply that for all @xmath256 sufficiently large @xmath486 where the last inequality follows , for all @xmath426 sufficiently small , from the definition of @xmath487 .
therefore , @xmath488 we next represent the measure @xmath489 as @xmath490 where @xmath491 are the positions of the particles alive at time @xmath492 . for the rest of the proof of the lemma we call the particle that is located at @xmath491 at time @xmath492
the @xmath244-th particle .
let @xmath493 be the measure describing the positions of the living descendents of the @xmath244-th particle at time @xmath494 and similarly to ( [ equt : nov3 ] ) we can write @xmath495 where @xmath496 is the position of the @xmath33-th descendent of the @xmath244-th particle at time @xmath494
. then we get that @xmath497 define @xmath498 then , @xmath499 hence , using lemma [ lem : nov1 ] in the first inequality , there exists @xmath477 sufficiently small such that @xmath500 where @xmath501 is the law of the standard brownian motion starting at @xmath15 . by taking the expectation
we conclude that for all @xmath256 sufficiently large , @xmath502 where @xmath476 , and we are done .
height5pt width5pt
[ [ proof - of - lemmalemnov4 ] ] proof of lemma [ lem : nov4 ] : + + + + + + + + + + + + + + + + + + + + + + + + + + fix @xmath503 sufficiently small such that @xmath504 .
let @xmath505 . if @xmath506 for each @xmath294 then the maximal displacement of the path of any particle on the time interval @xmath507 $ ] is bounded by @xmath508 hence by lemma [ lem : nov2 ] we get that for @xmath509 ,
@xmath510 now take @xmath511 sufficiently small so that for any @xmath512 @xmath513 and we are done .
height5pt width5pt [ lem:20_1 ] for any @xmath410
, there exists @xmath514 such that @xmath515 [ [ proof-11 ] ] proof : + + + + + + for any @xmath516 we have by lemma [ lem : nov4 ] that @xmath517 choose @xmath518 large enough so that @xmath519 , @xmath520 , and @xmath521 define @xmath522 then @xmath523 fix @xmath524 .
fix arbitrary @xmath525 and @xmath526 .
then there exists @xmath527 such that @xmath528 for @xmath529 let @xmath530 denote the smallest integer multiple of @xmath531 that is larger than @xmath109 and , with @xmath532 , let @xmath533 denote the largest integer multiple of @xmath531 that is smaller than @xmath306 .
let @xmath190 be any time such that @xmath534 .
then since @xmath535 and @xmath524 , we have by ( [ equt : nov5 ] ) and the continuity of @xmath536 that @xmath537 where the last inequality holds by ( [ 20_1 ] ) . by setting @xmath538
we are done .
width5pt the following corollary is immediate .
[ cor - immediate ] for any @xmath410 there exists @xmath539 such that , @xmath540 we have made all the preparation for the proof of the following lemma , concerning the tightness of the sequence @xmath155 . [ lem:07_06_1 ] the sequence of processes @xmath155 is @xmath158-tight in @xmath541 .
[ [ proof-12 ] ] proof : + + + + + + recall that the @xmath158-tightness of the sequence of the contour processes @xmath298 was proved in section [ sec - tcont ] ( see proposition [ prop : tight ] ) .
fix arbitrary @xmath542 and @xmath543 .
then for any @xmath539 , we have the following inclusion @xmath544 the @xmath158-tightness of the sequence @xmath155 now follows from this inclusion together with proposition [ prop : tight ] , corollary [ cor - immediate ] , and lemma [ lem:01_06_2]*(b)*. height5pt width5pt we next turn to the local time processes @xmath545 . [ lem - newtl ] the sequence of processes @xmath546 is @xmath158-tight in @xmath547 [ [ proof-13 ] ] proof : + + + + + + fix an arbitrary @xmath302 , and define @xmath548 with @xmath549 being as usual the corresponding measure .
note that since @xmath449 is arbitrary , it is enough to show the @xmath158-tightness of @xmath550 in @xmath551 and then the result follows immediately from lemma [ lem:01_06_2](b ) ( recall the properties of convergence in vague topology ) . since for each @xmath552 , the function @xmath553 is non - decreasing , to show the @xmath158-tightness of @xmath550 in @xmath551 , it is sufficient to prove the tightness of @xmath554 for each fixed @xmath243 .
that is , in view of lemma [ lem:01_06_2 ] , we need to prove that for any constant @xmath158 , @xmath555 the proof requires some care since introducing the time @xmath243 prevents one from directly exploiting martingale properties and the tightness results in @xcite .
we use the inverse local times @xmath556 to define the collection of processes @xmath557 note that @xmath558 represents the total mass of the branching process in random environment @xmath559 , defined by ( [ eq:31_12_01 ] ) , which starts at `` time '' @xmath560 , such that @xmath561 we also denote by @xmath562 the filtration generated by the process @xmath563 and its environment by time @xmath564 . on the event
@xmath565 we have , for any @xmath299 , @xmath566 } -\tilde \ell_{\tau^{n , i\delta}_{j\delta}}^{i\delta+v}| + \sup_{i\delta \leq t , j\delta \leq c_0 } \sup_{s\leq \delta } \bar{x}_s^{i , j,\delta}\nonumber\\ & & = : \sup_{i\delta \leq t , j\delta \leq c_0 } a_{i , j}+ \sup_{i\delta \leq t , j\delta \leq c_0 } b_{i , j}\ , .
\label{eq - technion1}\end{aligned}\ ] ] by the @xmath158-tightness of the sequence @xmath567 , see e.g. @xcite , theorem 4.2 ( proved there for the binary branching but valid , with similar proof , for the geometric case under consideration here ) , we have that for each fixed @xmath426 and each fixed @xmath568 , @xmath569 in particular , for any @xmath203 fixed , @xmath570 to control @xmath571 , we use the following lemma . [ lem - neqtech ] for some universal constant @xmath572 and all @xmath256 large , @xmath573 indeed , lemma [ lem - neqtech ] and chebychev s inequality imply that @xmath574 together with ( [ eq - technion5 ] ) , this yields the proof of lemma [ lem - newtl ] , once we complete the proof of lemma [ lem - neqtech ] . ' '' '' in the proof of lemma [ lem - neqtech ] we will frequently use the following lemma , whose immediate proof ( using iterations ) is omitted . [ recurs ] let @xmath575 and suppose @xmath576 satisfies the following inequalities @xmath577 then there exists @xmath578 such that for any @xmath579 $ ] @xmath580 [ [ proof - of - lemma - lem - neqtech ] ] proof of lemma [ lem - neqtech ] : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the argument uses computations similar to those in section [ sec - asymp ] . throughout the proof , @xmath581 denotes a constant whose value may change from line to line , but is independent of @xmath256 or @xmath426 .
note that the estimates on @xmath558 that we get throughout the proof below are uniform in @xmath582 and thus we may and will just consider @xmath583 and write @xmath584 and @xmath585 .
note that @xmath586 is the local time at level @xmath190 accumulated by the random walk during its first @xmath587 excursions from @xmath9 .
we have the representation @xmath588 where the @xmath589 is the number of offspring of the @xmath33-th particle at time @xmath590 .
recall that @xmath591 are conditionally indpendent given @xmath592 , and for each @xmath33 , @xmath593 is geometrically distributed with parameter @xmath594 . here with some abuse of notation , @xmath595 @xmath12 is as in section [ 1.1 ] , and @xmath596 is the position of @xmath33-th particle at time @xmath232 . note that by ( [ eqm32 ] ) and our moment assumptions on @xmath12 we have that @xmath597 because the mean of @xmath593 is close to @xmath420 , the sequence @xmath598 is almost a martingale . to make it into a martingale ,
introduce the variables , @xmath599 , @xmath600 note that @xmath601 on the other hand , @xmath602 is a discrete martingale , and hence by the doob - burkholder - gundy inequality , we have that @xmath603 where @xmath604 we prepare next some estimates . first recall ( [ eqm32 ] ) , our moment assumptions on @xmath12 and its covariance structure to get the following bound on the correlation between the @xmath605 : @xmath606| \leq \bar c / n\,,\;\;\forall k\not= k'.\ ] ]
then we easily get , @xmath607\nonumber\\ & \leq & \bar c\frac{m_i}{n}+\bar c \frac { m_i^2}{n}\,,\end{aligned}\ ] ] note that @xmath608 , and hence to control the right side of ( [ eq - wis11 ] ) we need to bound @xmath609 .
@xmath610 is a martingale and hence with @xmath611 we use ( [ eq - wis11 ] ) to get @xmath612 by lemma [ recurs ] we get @xmath613 now recall again that @xmath614 and use the above and ( [ eq - wis11 ] ) to obtain that @xmath615 a similar computation , using remark [ rem - new ] , gives @xmath616 with @xmath617 one then obtains the recursions @xmath618 for @xmath256 sufficiently large ( @xmath619 ) , and therefore by lemma [ recurs ] we have @xmath620
repeating this computation for the fourth moment , one obtains that with @xmath621 , @xmath622 for all @xmath256 sufficiently large . substituting ( [ eq - wis11 ] ) into ( [ eq - wis1 ] ) and using the last estimates , one gets @xmath623 for all @xmath256 sufficiently large . since , by ( [ eqm33 ] ) , @xmath624 this completes the proof of lemma [ lem - neqtech ] . ' '' '' [ cor:4 ] @xmath450 is @xmath158-tight in @xmath625 .
[ [ proof-14 ] ] proof : + + + + + + immediately from lemma [ lem - newtl ] and lemma [ lem:07_06_1 ] . ' '' '' in what follows let @xmath626 be a limiting point of @xmath627 . to simplify the notation we omit subsequences and
simply assume that @xmath628 converges to @xmath626 .
we also switch ( by skorohod s theorem ) to some probability space where the convergence holds a.s .. recall again that we write @xmath629 and @xmath630 for @xmath631)$ ] and @xmath632)$ ] respectively .
[ lem - loctime ] @xmath455 is the local time of @xmath162 .
[ [ proof-15 ] ] proof : + + + + + + first note that by properties of weak convergence of measures , for any @xmath108 @xmath633 for any point of continuity of function @xmath634 .
however by a limiting argument and the convergence of @xmath93 to @xmath162 , it is easy to derive that if @xmath635 then @xmath190 is a point of continuity of @xmath634 . therefore , for all @xmath636 such that @xmath637 , ( [ 01_06_3 ] ) follows .
note that @xmath638 also for any @xmath108 and @xmath203 we have @xmath639 since @xmath640 is tight and @xmath426 was arbitrary we can make the left side arbitrarily small by taking @xmath203 sufficiently small with probability as close to @xmath420 as we wish uniformly in @xmath256 .
this , by a standard argument , that also uses the convergence of @xmath298 , implies that @xmath641 for any @xmath642 . on the other hand @xmath643 where the last convergence follows by convergence of @xmath644 at all the points @xmath645 such that @xmath646 ( there is just one level @xmath647 such that @xmath648 ) .
this and ( [ 01_06_4 ] ) yield @xmath649 for all @xmath650 , and hence @xmath651 is indeed the local time of @xmath162 , for any @xmath99 . ' '' ''
the above lemma and corollary [ cor:4 ] finish the proof of proposition [ prop:3 ] .
the next two lemmas are essential for the proof of the charaterization of the limit points " part of theorem [ thm:1 ] .
first we prove the continuity of the local time at the level zero .
[ lem:29_12_2 ] @xmath652 is continuous .
[ [ proof-16 ] ] proof : + + + + + + it is enough to show that for arbitrary @xmath302 , @xmath653 is @xmath158-tight in @xmath301 , that is , for any @xmath410 @xmath654 suppose ( [ eq:27_1 ] ) does not hold , that is , there exist @xmath655 , such that for all @xmath203 @xmath656 fix such @xmath655 ; we have the inclusion @xmath657 since @xmath658 are identically distributed we get @xmath659 by lemma [ lem:01_06_2](a ) , we can choose @xmath426 sufficiently small such that @xmath660 for all @xmath256 sufficiently large , and hence @xmath661 and we get a contradiction with ( [ eq:27_2 ] ) . ' '' '' [ lem:28_12_1 ] for any fixed @xmath164 , @xmath408 equals , with probability one , to the value of the inverse function of @xmath662 at @xmath647 , that is , @xmath663 [ [ proof-17 ] ] proof : + + + + + + recall that we assume that we are considering the probability space where @xmath664 in @xmath665 , @xmath666-a.s .. moreover we know that for any fixed @xmath647 , @xmath667 is continuous at the point @xmath647 .
this , by properties of convergence in @xmath668 , implies that for any fixed @xmath647 , @xmath669 , @xmath666-a.s .. fix arbitrary @xmath670 .
then , by definition of the local time , we get , @xmath671 since @xmath672 converges to the continous limit , the convergence is uniform on the compacts .
this and the convergence @xmath673 imply , that by passing to the limit in ( [ eq:29_12_1 ] ) we get @xmath674 and hence @xmath675 similarly we can show that @xmath676 since @xmath426 was arbitrary , and by the continuity of @xmath677 at @xmath449 ( see lemma [ lem:01_06_2]*(d ) * ) we get @xmath678 and we are done . ' '' '' [ lem:27_12_3 ] for any @xmath679 and fixed @xmath302 , @xmath680 as@xmath681 where @xmath682 [ [ proof-18 ] ] proof : + + + + + + @xmath683 , where by lemma [ lem:28_12_1
] @xmath684 is defined by ( [ eq:29_12_4 ] ) . moreover , by lemma [ lem:29_12_2 ] , @xmath662 is continuous at @xmath685 , therefore by elementary properties of weak convergence , for any continnuous function @xmath210 @xmath686 now the result for @xmath1 , follows by uniform on the compacts convergence of @xmath687 to @xmath688 .
the convergence of the integral for @xmath689 follows immediately since , by the continuity of @xmath162 , the @xmath690 does not charge the point @xmath691 for every @xmath689 . ' '' '' [ [ proof - of - theorem - thm1 ] ] proof of theorem [ thm:1 ] : + + + + + + + + + + + + + + + + + + + + + + + + + the tightness statement was proved in proposition [ prop:3 ] .
to finish the proof we need to derive the characterization of the limit points .
fix arbitrary @xmath302 and let @xmath692 be the measure - valued process defined as in ( [ eq : oc5_1 ] ) , that is , @xmath693,\end{aligned}\ ] ] for all @xmath114 .
let @xmath694 be an arbirary limit point of + @xmath695 .
fix arbitrary @xmath696 .
as we have mentioned already , due to results in @xcite , the sequence of process @xmath405 converges weakly in @xmath697 $ ] to the process @xmath698 $ ] satisfying the martingale problem ( [ mp-1]-[mp-2 ] ) on @xmath166 $ ] , with @xmath699 , and hence the left hand side of ( [ 20_3 ] ) converges to @xmath700 for any @xmath701 $ ] . as for the right hand side of ( [ 20_3 ] ) , due to proposition [ prop:3 ] and lemma [ lem:27_12_3 ] it converges , along an appropriate subsequence , to @xmath702 for @xmath703 $ ] , where @xmath455 is the local time @xmath162 .
this gives us ( [ 20_2 ] ) for any @xmath696 .
the extension of the equality to any @xmath114 is trivial . ' '' ''
the proof of the result is based on convergence of approximations . for simplicity , as before
, we assume that @xmath704 a.s .
( based on proposition [ prop:3 ] we can always get it by skorohod s theorem via an appropriate subsequence ) .
further , we localize the snake @xmath705 to live in a compact , and then it is not hard to check that for @xmath256 large , the truncation in the definition of @xmath706 can be ignored .
thus , we assume in the sequel that @xmath707 on the level of the @xmath256th approximation we will be dealing with the following approximating functional : @xmath708 note that @xmath709 further , if @xmath710 , then @xmath711 where @xmath94 is a brownian path independent of @xmath712 . let @xmath713 define @xmath714 then by the standard decomposition of @xmath715 we get that @xmath716 where @xmath717 is the @xmath718-martingale given by @xmath719 and @xmath720 we first study the limiting behavior of @xmath721 .
[ lem:5.1 ] @xmath722 [ [ proof-19 ] ] proof : + + + + + + using @xmath723 to denote expectation with respect to the brownian path @xmath724 , we have @xmath725\ ) } \bigg|y^n_{(k+1)/n^2 } > y^n_{k / n^2}\,,\ \cf_k\right)\,.\end{aligned}\ ] ] therefore , @xmath726 \frac{1}{n}e^{-\frac{1}{\sqrt{n}}\sum_{l'=1}^{ny^n_{k / n^2}-1}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k})}\\ & & \mbox{}+\left[\frac{1}{2}+\frac{1}{4\sqrt{n}}\xi_{y^n_{k / n^2}}\left ( \hat{\ww}^n_{n^{-2}k}\right ) \right ] \ee_{\eta}\(\frac{1}{n}e^{-\frac{1}{\sqrt{n}}\sum_{l'=1}^{ny^n_{k / n^2}}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k}+\eta_{1/n})}\)\\ & & \mbox{}+\frac{1}{n}{\bf 1}_{\{y^n_{k / n^2}=0\ } } -{\bf 1}_{\ { y^n_{k / n^2}=k_1\}}\frac{1}{n}e^{-\frac{1}{\sqrt{n}}\sum_{l'=1}^{nk_1 - 1}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k } ) } \\ & = & \frac{1}{2n}\left[\ee_{\eta}\(e^{-\frac{1}{\sqrt{n}}\sum_{l'=1}^{ny^n_{k / n^2}}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k}+\eta_{1/n})}\)- e^{-\frac{1}{\sqrt{n}}\sum_{l'=1}^{ny^n_{k / n^2}-1 } \xi_{l'}(\hat{\ww}^{n}_{n^{-2}k})}\right]\\ & & \mbox{}+\frac{1}{4n^{3/2}}\xi_{y^n_{k / n^2}}\left ( \hat\ww^n_{n^{-2}k}\right ) \left[e^{-\frac{1}{\sqrt{n } } \sum_{l'=1}^{ny^n_{k / n^2}-1}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k } ) } \right.\\ & & \left . \hspace*{2cm}\mbox { } + \ee_{\eta}\(e^{-\frac{1}{\sqrt{n}}\sum_{l'=1}^{ny^n_{k / n^2}}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k}+\eta_{1/n})}\ ) \right ] \\ & & \mbox{}+\frac{1}{n}\cdot { \bf 1}_{\{y^n_{k / n^2}=0\ } } -\frac{1}{n}\cdot { \bf 1}_{\{y^n_{k / n^2}=k_1\}}e^{-\frac{1}{\sqrt{n}}\sum_{l'=1}^{nk_1 - 1}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k } ) } \\ & = & \frac{1}{2n}\left[\ee_{\eta}\(e^{-b_{y^n_{k / n^2}}(\hat{\ww}^{n}_{n^{-2}k}+\eta_{1/n})}\)- e^{-b_{y^n_{k / n^2}-1/n}(\hat{\ww}^{n}_{n^{-2}k})}\right]\\ & & \mbox{}+\frac{1}{4n^{3/2}}\xi_{y^n_{k / n^2}}\left ( \hat{\ww}_{n^{-2}k}\right ) \left(e^{-b_{y^n_{k
/ n^2}-1/n}(\hat{\ww}^{n}_{n^{-2}k})}+ \ee_{\eta}\(e^{-b_{y^n_{k / n^2}}(\hat{\ww}^{n}_{n^{-2}k}+\eta_{1/n})}\ ) \right ) \\ & & \mbox{}+\frac{1}{n}\cdot { \bf 1}_{\{y^n_{k / n^2}=0\ } } -\frac{1}{n}\cdot { \bf 1}_{\{y^n_{(k-1)/n^2}=k_1 - 1,y^n_{k / n^2}=k_1\ } } e^{-b_{k_1 - 1/n}(\hat{\ww}^{n}_{n^{-2}k})}\\ & = & i_{1,n , k}+i_{2,n , k}+i_{3,n , k}-i_{4,k , n}\,,\end{aligned}\ ] ] where we also used the definition of @xmath183 .
we begin with an estimate of @xmath727 .
set @xmath728\,.\end{aligned}\ ] ] by it s formula we get @xmath729 the first term at the right side above can be further decomposed as @xmath730\,,\end{aligned}\ ] ] where @xmath731 is some point in @xmath732 $ ] .
we get @xmath733\right ) \\ & & \mbox{}+ \left.\ee_{\eta}\left(\int_{0}^{1/n}r_{n , k}(s ) ds\right)\right ] .
\end{aligned}\ ] ] to handle @xmath734 , note that @xmath735\right ) + \left.\ee_{\eta}\left(\int_0^{1/n } r_{n , k}(s)\,ds\right)\right].\end{aligned}\ ] ] all together we get @xmath736 from this it follows that for any @xmath689 @xmath737 where , due to the hlder continuity of the gaussian field , the second and third terms on the right side of the first equality converge to @xmath9 ( once @xmath738 has been localized in a compact with high probability ) , and the first term converges to the first term on the right side of ( [ eq:5.2 ] ) .
now we will treat @xmath739 and @xmath740 . by definition of the approximate local time @xmath741 we get @xmath742
then pass to the limit , use the uniform on compacts convergence of @xmath743 and @xmath81 to @xmath455 and @xmath147 , and the continuity of @xmath183 to get that @xmath744 as @xmath97 .
thus , we obtain the second and the third terms in ( [ eq:5.2 ] ) . ' '' '' define the bracket process for the martingale @xmath201 : @xmath745 then we have [ lem:5.2 ] @xmath746 [ [ proof-20 ] ] proof : + + + + + + it is easy to check that for any @xmath747 , @xmath748 by lemma [ lem:5.1 ] we know that as @xmath97 @xmath749 which is a process of bounded variation . from this
it is easy to deduce that @xmath750 as @xmath224 .
hence it is enough to consider the limiting behavior of @xmath751 by repeating the argument in the proof of lemma [ lem:5.1 ] we get @xmath752 \frac{1}{n^2}e^{-\frac{2}{\sqrt{n}}\sum_{l'=1}^{ny^n_{k / n^2}-1}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k})}\\ & & \mbox{}+\left[\frac{1}{2}+\frac{1}{4\sqrt{n}}\xi_{y^n_{k / n^2}}\left ( \hat\ww^n_{n^{-2}k}\right ) \right ] \ee_{\eta}\(\frac{1}{n^2}e^{-\frac{2}{\sqrt{n}}\sum_{l'=1}^{ny^n_{k / n^2}}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k}+\eta_{1/n})}\)\\ & & \mbox{}+\frac{1}{n^2}1_{y^n_{k / n^2}=0 } + 1_{y^n_{k / n^2}=k_1}\frac{1}{n^2}e^{-\frac{2}{\sqrt{n}}\sum_{l'=1}^{nk_1 - 1}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k})}\,.\end{aligned}\ ] ] therefore , @xmath753\\ & & \mbox{}+\frac{1}{4n^{5/2}}\xi_{y^n_{k / n^2}}\left ( \hat\ww^n_{n^{-2}k}\right ) \left[-e^{-\frac{2}{\sqrt{n}}\sum_{l'=1}^{ny^n_{k
/ n^2}-1}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k } ) } \right.\\ & & \left.\;\;\;\;\ ; \mbox{}+ \ee_{\eta}\(e^{-\frac{2}{\sqrt{n}}\sum_{l'=1}^{ny^n_{k / n^2}}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k}+\eta_{1/n})}\ ) \right ] \\ & & \mbox{}+\frac{1}{n^2}\cdot { \bf 1}_{\{y^n_{k / n^2}=0\ } } + \frac{1}{n^2}\cdot { \bf 1}_{\{y^n_{k / n^2}=k_1\ } } e^{-\frac{2}{\sqrt{n}}\sum_{l'=1}^{nk_1 - 1}\xi_{l'}(\hat{\ww}^{n}_{n^{-2}k } ) } \\ & = & \frac{1}{2n^2}\left(\ee_{\eta}\(e^{-2b_{y^n_{k / n^2}}(\hat{\ww}^{n}_{n^{-2}k}+\eta_{1/n})}\)+
e^{-2b_{y^n_{k / n^2}-1/n}(\hat{\ww}^{n}_{n^{-2}k})}\right)\\ & & \mbox{}+\frac{1}{4n^{5/2}}\xi_{y^n_{k / n^2}}\left ( \hat{\ww}_{n^{-2}k}\right ) \left(-e^{-2b_{y^n_{k
/ n^2}-1/n}(\hat{\ww}^{n}_{n^{-2}k})}+ \ee_{\eta}\(e^{-2b_{y^n_{k / n^2}}(\hat{\ww}^{n}_{n^{-2}k}+\eta_{1/n})}\ ) \right ) \\ & & \mbox{}+\frac{1}{n^2}\cdot { \bf 1}_{\{y^n_{k / n^2}=0\ } } + \frac{1}{n^2}\cdot { \bf 1}_{\{y^n_{k / n^2}=k_1\ } } e^{-2b_{k_1 - 1/n}(\hat{\ww}^{n}_{n^{-2}k})}\\ & = & j_{1,n , k}+j_{2,n , k}+j_{3,n , k}+j_{4,n , k}\,.\end{aligned}\ ] ] using the bounds from the proof of lemma [ lem:5.1 ] it is easy to see that @xmath754 as @xmath224 .
as for @xmath755 , again using the convergence of @xmath756 and the continuity of @xmath183 , it is easy to see that @xmath757 and we are done . ' '' '' [ cor:5.1 ] as @xmath224 , @xmath201 converges to a continuous local martingale @xmath758 such that @xmath759 [ [ proof-21 ] ] proof : + + + + + + the continuity of @xmath758 is immediate from the continuity of the limiting process @xmath162 and lemma [ lem:5.1 ] .
the rest is immediate from lemma [ lem:5.2 ] . ' '' '' [ cor:5.2 ] there exists a brownian motion @xmath184 such that @xmath760 [ [ proof-22 ] ] proof : + + + + + + immediate from the previous corollary .
[ [ proof - of - theoremthr5.1 ] ] proof of theorem [ thr:5.1 ] : + + + + + + + + + + + + + + + + + + + + + + + + + + + immediate from lemma [ lem:5.1 ] , corollary [ cor:5.1 ] and corollary [ cor:5.2 ] . ' '' '' finally , we describe the snake process when @xmath189 is constant .
the description for the general case , more specifically , the uniqueness of the solution for the martingale problem ( [ eq:5.1 ] ) remains a challenging _ open _ problem .
when @xmath189 is constant , say @xmath761 , we have that @xmath762 is a brownian motion with constant drift @xmath763 .
it follows from the martingale problem ( [ eq:5.1 ] ) that @xmath764 therefore , @xmath765 is the brox diffusion reflected at @xmath9 and @xmath145 ( see the appendix for a description when @xmath192 ) .
next , we consider the conditional ( given the lifetime process ) path process .
let @xmath766 be an element in @xmath68 .
fix @xmath767 $ ] and @xmath768 .
similar to legall ( @xcite , p54 ) , we define @xmath769 as the unique probability measure on @xmath68 such that + ( i ) @xmath770 a.s .
+ ( ii ) @xmath771 for all @xmath772 , @xmath769 a.s .
+ ( iii ) under @xmath769 , @xmath773)$ ] is a brownian motion .
denote the time set @xmath774 . from the construction of the discrete snake
, it follows that @xmath775 is a conditional ( given @xmath93 ) markov chain with transition probability @xmath776 where @xmath777\cap q_n\}$ ] .
taking @xmath778 , we see that the limit @xmath779 is a conditional ( given @xmath162 ) markov process with transition probability @xmath780 where @xmath781\}$ ] .
namely , it has the same conditional law as legall s brownian snake .
we provide in this appendix a short , direct proof of corollary [ cor - brox ] that bypasses the study of the branching process , relying instead on an embedding of a random walk in random environment ( rwre ) into a diffusion in random environment , in the spirit of @xcite .
for backround on brownian motion in random environments we refer to @xcite , @xcite , @xcite and to the nice overview in @xcite .
background for rwre can be found in @xcite .
recall that a brownian motion in random environment ( bmre ) is a process @xmath782 given by @xmath783 where @xmath784 is a brownian motion and @xmath785 is called the random potential . when @xmath785 is itself a brownian motion independent of @xmath786 , this ( formal ) process is the brox diffusion @xcite .
we need to consider reflecting bmre s .
let @xmath487 be the periodic function with period @xmath787 and @xmath788 for @xmath789 .
let @xmath785 be a brownian motion on @xmath790 $ ] and set @xmath791 for @xmath792 .
set formally @xmath793 ( in case @xmath785 is not smooth , a precise meaning is given to ( [ eq - oo1 ] ) by the procedure described in ( * ? ? ?
* section 2 ) ) .
let @xmath794 .
a _ formal _ application of the it - tanaka formula yields @xmath795 where @xmath796 is a brownian motion . to justify ( [ isitfinal ] ) ,
one argues as follows .
first , an application of ito s formula for dirichlet processes , see e.g. @xcite , gives that for any @xmath189 which is twice differentiable , and with @xmath797 , @xmath798 now note that , by definition of the local time as the occupation time density , the local times of @xmath799 and @xmath162 at levels @xmath9 and @xmath145 are equal up to multiplicative constant @xmath800
. therefore a standard approximation of @xmath487 by smooth functions @xmath189 , together with ( [ 1601 ] ) , yields ( [ isitfinal ] ) , provided that the local time @xmath801 of @xmath802 is jointly continuous in @xmath243 and @xmath15 , the latter at @xmath803 and @xmath804 .
however , @xmath801 is a continuous transformation of the local time of the brownian motion @xmath805 ( see e.g. equation ( 10 ) in @xcite for an explicit formula which holds for any environment not necessarily for the two sided white noise ) , and thus is jointly continuous in its arguments .
this yields ( [ isitfinal ] ) . therefore , @xmath806 is a reflecting ( at @xmath9 and @xmath145 ) brox diffusion .
let the environment be given by a family @xmath807 of independent random variables with mean @xmath9 and variance 1 .
we further assume that @xmath808 .
define the potential @xmath809 on @xmath104 by @xmath810}_{i=1}\log\frac{\frac{1}{2}-\frac{1}{4\sqrt{n}}\xi^n(i ) } { \frac{1}{2}+\frac{1}{4\sqrt{n}}\xi^n(i)},\ ] ] and set @xmath811 and let @xmath812 be the bmre with potential @xmath813 .
set @xmath814 .
define the stopping times @xmath815 and @xmath816 [ rem:12_01 ] note that @xmath822 is a sequence of reflecting ( at @xmath9 and @xmath823 ) rwre such that @xmath824 and hence by the continuity of the function @xmath487 and the discussion in the beginning of the appendix , in order to prove corollary [ cor - brox ] it is sufficient to prove proposition [ prop:12_01 ] .
let @xmath826 as @xmath827 is the bmre with potential @xmath813 , it is well - known ( see ( 2.3 ) in @xcite ) that @xmath828 is a local martingale with quadratic variation @xmath829 such that @xmath830 we now rescale .
let @xmath831 then @xmath832 thus @xmath833 is a local martingale with quadratic variation process @xmath834 .
thus , @xmath835 therefore ( see again ( 2.3 ) and ( 2.5 ) in @xcite ) , @xmath200 is the bmre with potential @xmath825 . as was proved in the previous subsection ( see lemma [ lem:12_01_2 ] ) , the scaled rwre is related to bmre by @xmath836}=z^n({{\sigma}^n_{[n^2t]}}).\end{aligned}\ ] ] in this section , we first prove that @xmath837}\to t,\;\ ; { \rm as}\ ; n\rightarrow\infty , \end{aligned}\ ] ] by the strong law of large numbers . then , we prove that the scaled potential for @xmath200 converges to @xmath838 , and hence @xmath200 converges to a bmre with potential @xmath838 .
this by ( [ eq:12_01_2 ] ) and ( [ eq:12_01_3 ] ) will provide the proof of proposition [ prop:12_01 ] . by proposition 3.2 in @xcite ( or a direct computation involving a time change ) , we see that @xmath840 , @xmath841 , are i.i.d . with the same distribution as @xmath842 where @xmath843 is a standard brownian motion .
note that @xmath844 . by the strong law of large numbers
, we get that @xmath839}=t\frac{1}{n^2t}\sum^{[n^2t]}_{i=1 } \theta_i\to t\ ] ] uniformly on compacts .
first we consider the weak convergence of @xmath847 .
note that @xmath848}_{i=1}\frac{1}{\sqrt{n}}\xi^n(i)+o(1)\equiv m^n_x + o(1).\ ] ] regarding @xmath15 as the time - parameter , @xmath849 is a martingale with predictable quadratic variation process @xmath850}_{i=1}\ee\(\frac{1}{\sqrt{n}}\xi^n(i)\)^2\to x\ ] ] uniformly on the compacts .
thus , by theorem 4.13 ( @xcite , p358 ) , @xmath201 converges weakly in @xmath301 to a brownian motion @xmath851 . by switching to another probability space if necessary , we may and will assume that all weak convergences hold almost surely
. then we get , @xmath852 , a.s .. note that by the continuity of @xmath487 , we immediately get that @xmath853 uniformly on the compacts of @xmath104 .
note that ( see ( 2.6 ) in @xcite ) , @xmath854 where @xmath855 @xmath856 and @xmath857 is a brownian motion . since @xmath857 trivially converges weakly to the brownian motion @xmath843
, we assume as before that the convergence holds a.s .. then we have @xmath858 and @xmath859 note that all the convergence above are a.s . and uniform on compacts .
we see that @xmath860 now proposition [ prop:12_01 ] follows from lemmas [ lem:12_01 ] , [ lem - broxmain ] , and ( [ eq:12_01_2 ] ) .
then as we have mentioned already in remark [ rem:12_01 ] , corollary [ cor - brox ] follows immediately from proposition [ prop:12_01 ] .
le gall ( 1996 ) .
_ superprocesses , brownian snakes and partial differential equations_. lecture notes from the 11th winter school on stochastic processes , sigmundsburg , prpublication 337 du laboratoire de probabilits , universit paris vi .
e. perkins ( 2002 ) .
watanabe superprocesses and measure - valued diffusions . _ lectures on probability theory and statistics , saint - flour 1999 _ , lecture notes in mathematics , * 1781 * , springer , berlin , 132329 . | we consider ( discrete time ) branching particles in a random environment which is i.i.d . in time and possibly spatially correlated . we prove a representation of the limit process by means of a brownian snake in random environment .
[ section ] [ thm]example [ thm]corollary [ thm]notation [ thm]notations [ thm]definition [ thm]lemma [ thm]claim [ thm]meta - theorem [ thm]proposition [ thm]remark [ thm]conjecture [ thm]remarks @xmath0 ) = msbm10==msbm7 = # 1#1 a |
in many fields of astronomy , image archives are of increasing importance .
for example , since 2005 , more than 50% of hst papers have been based on archival data rather than pi data .
archival images have become increasingly useful to extra - galactic and stellar astronomers in the last few years but , until now , solar system researchers have been at a disadvantage in this respect . while regular archive searches can find images at a fixed location , they can not find images of moving targets .
this is unfortunate , because it could be argued that archival data is potentially more useful to solar system studies than extra - galactic or stellar astronomy .
the full scientific potential derived from the discovery of small solar system bodies can not be fully realized until precise orbital parameters for those objects can be determined .
often , an object is detected but an orbit can not be determined until a significant amount of time has passed . if archival ( precovery ) images of the object exist , one can determine the orbital parameters immediately .
the solar system object search ( ssos ) at the canadian astronomy data centre ( cadc ) allows users to search for images of moving objects .
the user enters either a set of observations , an object name , orbital elements or a full ephemeris .
ssos generates an ephemeris and returns a list of all matching images .
before any queries can be made , the ssos image database must be populated .
this is done by going to various telescope archives and harvesting the metadata describing each image taken by the telescope . for each image
ssos stores the following information : midpoint of exposure time , ra and dec of the image center , and the extent of the image in ra and dec . in addition , ssos stores a bounding box in time , ra and dec .
the bounding box is in integer days ( mjd ) and integer degrees for ra and dec .
obtaining the metadata for images stored at the cadc , where ssos is based , is relatively easy .
the existing databases describing each image archive are queried directly and the relevant parameters are ingested into the ssos database . to date , the archives of the following telescopes have been harvested from the cadc : cfht ( megacam and wircam ) , gemini ( gmos ) , and hst ( wfpc , acs , and wfc3 ) .
offsite archives must be `` scraped '' over the web .
this can take many forms depending on the archive .
the subaru suprimecam image lists are available as simple ascii text files .
the eso archive can be queried repeatedly by date , eventually returning all the observations made by those telescopes . to date , the archives of the following telescopes have been harvested : aat ( wfi ) , eso - lasilla 2.2 m ( wfi ) , eso - ntt ( efosc , emmi , sofi , susi / susi2 ) , eso - vista ( vircam ) , vlt ( fors1 , fors2 , hawki , naos - conica , isaac and vimos ) , and subaru ( suprimecam ) . currently , there are 1.6 million images in the ssos database
. figure [ fig : coverage ] shows the area of the sky covered by ssos .
when arriving at the solar system object search tool website , users have four ways to search for images . in each case
, ssos converts the user s input into an ephemeris .
the four methods of input and conversion are detailed in the following four subsections : in this input method , the user enters a series of observations in mpc format .
ssos then uses these observations to determine an orbit and generate an ephemeris from that orbit
. the user can select one of two orbit fitting routines : the orbit fitting code of @xcite has been set up to automatically convert the observations into orbital parameters ( fitradec ) and use those parameters to produce an ephemeris ( predict ) .
ssos also provides the new object ephemeris generator from the minor planet center as an alternative .
if a user selects this option , the ssos queries the mpc service automatically .
the mpc fits a visl orbit to the observations and returns an ephemeris based on this orbit .
this method is slower than the bernstein & khushalani fitting because it requires ssos to make queries to an external service .
the ephemeris is generated at intervals of 24 hours .
mauna kea ( observatory code 568 ) is used as the observing site . in this input method
, the user enters the name of an object .
ssos then forwards that name to one of two services , either the lowell observatory asteroid ephemeris generator or the minor planet and comet ephemeris service at the minor planet center .
these services query their databases for an object matching the name , make the appropriate orbital calculations and return an ephemeris to ssos .
in addition to using these two offsite services , ssos can also generate an ephemeris locally .
ssos maintains a regularly updated copy of the mpc orbital element database .
when a user enters an object name , the local version of this database is queried and the orbital elements are passed to the program orbfit from the orbfit software package @xcite , which generates an ephemeris . as with the search by arc option ,
the ephemeris is generated at 24 hour intervals and mauna kea is used as the observing site . in this case
, the user enters the orbital elements of an object : epoch , semi - major axis , eccentricity , inclination , longitude of the ascending node , argument of perihelion and mean anomaly .
these orbital elements are used as input to the program orbfit which returns an ephemeris , again at 24 hour intervals and using mauna kea as the observing location .
this method allows the user complete control over the ephemeris .
the user enters a series of times and object positions .
users can `` cut and paste '' text into the service .
this method is useful if the user has any concerns about the positional accuracy of any of the previous methods .
for example , the object in question might be near enough to the earth that the parallax will significantly affect the objects positions .
alternatively , the object s apparent motion might be irregular enough that the linear daily interpolation scheme is insufficiently accurate .
once an ephemeris has been generated by one of the above methods , the next step is to match that ephemeris to the database of observations .
the ephemeris can be thought of as a series of line segments in 3 dimensions : time , ra and dec .
ssos builds an integer - valued bounding box ( in both time and position ) around each segment .
the ephemeris is saved as a temporary table with both the integer bounding boxes and the exact ( floating point ) values .
when the object moves across the first point of aries , two rows are added , one each describing the position of the object on either side of the celestial meridian .
ssos then matches this temporary table to its observation table . for speed ,
the integer bounding boxes are matched first .
if a match is possible ( _ i.e. _ , if the bounding boxes overlap ) ssos does a linear interpolation to determine the position of the object at the time of the exposure .
a match occurs if this position is within the footprint of the image .
the key to keeping the queries reasonably fast are the integer bounding boxes and the linear interpolation .
doing a full orbital prediction for each of the images is not feasible .
this is sufficiently accurate for the majority of queries , where the object either moves slowly or in a fairly straight line .
a typical 20 year ephemeris can be matched to 1.6 million images in less than a second .
ssos then returns a list of matching images to the user .
where possible , direct links to the images are provided ; otherwise , ssos provides links to pages where the images can be requested . on average , a search will return 20 images of a given asteroid , but this number can range from 0 ( no hits ) to several hundred . | while regular archive searches can find images at a fixed location , they can not find images of moving targets such as asteroids . the solar system object search ( ssos ) at the canadian astronomy data centre allows users to search for images of moving objects .
ssos accepts as input either a list of observations , an object designation , a set of orbital elements , or a user - generated ephemeris for an object .
it then searches for observations of that object over a range of dates .
the user is then presented with a list of images containing that object from a variety of archives . initially created to search the cfht megacam archive , ssos has been extended to other telescope archives including gemini , subaru / suprimecam , hst , and several eso instruments for a total of 1.6 million images .
the ssos tool is located on the web at : http://www.cadc.hia.nrc.gc.ca/ssos |
hadron masses reveal key aspects of qcd .
the light pseudoscalar meson masses reveal the spontaneous breaking of chiral symmetry as well as its explicit breaking by the light quark masses , and in a rather direct manner they permit to extract the ratios of these quark masses which are fundamental parameters of qcd .
they also give access to the electromagnetic contributions to the masses , for instance , in the mass difference between charged and neutral pions . in the framework of chiral perturbation theory , a rather accurate understanding of the effects of light quark masses and electromagnetic corrections on the pseudoscalar octet has been achieved , and to a lesser extent in baryons as well .
heavy mesons represent another kind of hadronic system where one can arrive at a good determination of the various effects that determine their masses .
also , to a first degree of approximation in the heavy quark expansion , the strong hyperfine effects , which involve the heavy quark spin , serve to determine the ratio @xmath4 , which is another fundamental input in qcd . in this work , we show that the current knowledge of the heavy meson masses allows for a quantitative determination of the different effects that contribute to the differences of heavy meson masses .
the approach followed here is similar to the one given by rosner and wise @xcite , and the improvement in the results is possible thanks to the better empirical accuracy in the heavy meson masses and also in the better knowledge of the heavy quark masses . in the limit of infinite heavy quark masses and of light quark @xmath5 symmetry ,
heavy mesons fill multiplets of @xmath6 , where @xmath7 is the number of heavy quarks , @xmath8 is the rotation group associated with the light degrees of freedom of the heavy meson . in particular , the ground state mesons , namely @xmath0 , @xmath1 , @xmath2 and @xmath3 , fill the multiplet @xmath9 of that symmetry .
the symmetry is broken by several effects : * the finite masses of the heavy _
c _ and _ b _ quarks break @xmath10 , where @xmath11 is the rotation group is associated with the spin of the meson .
* the light quark masses break @xmath12 . *
the em interactions break all symmetry subgroups down to @xmath13 .
the analysis in this work has the objective of sorting out these three sources of symmetry breaking from the current knowledge of the heavy meson masses .
the analysis only involves meson mass differences .
those in the @xmath0 mesons are very well known with errors smaller than 2.5% @xcite . in the @xmath2 mesons ,
not all mass splittings are established ; the mass differences @xmath14 and @xmath15 are not separately known and , in addition , the errors are significantly larger than in the d - system @xcite . nonetheless , the available information is sufficient for the analysis to lead to significant conclusions .
the analysis is based on the mass formuli that result from the expansions in @xmath16 , where @xmath17 is the heavy quark mass @xmath18 , in @xmath19 and @xmath20 , and in the fine structure constant @xmath21 .
these mass formuli are similar to the ones given long ago by rosner and wise @xcite , except that the qcd running of some of the parameters are taken into account . using the notation @xmath22 , etc .
, they read as follows : @xmath23 where @xmath24 and @xmath25 denote respectively the light and heavy quark flavors of the heavy meson , the label @xmath26 is used for the vector mesons , and @xmath27 and @xmath28 are the respectively the light and heavy quark charges . the first term in the mass formula is made out of the contributions in the limit @xmath29 and @xmath30 plus @xmath5 singlet contributions by light quark masses , _
i.e. _ given in terms of @xmath31 , and by electromagnetism .
these @xmath5 symmetric contributions are simply absorbed into the coefficients that determine @xmath32 in eqn.(2 ) below .
our analysis can not determine those contributions separately .
the second term in the mass formula provides the @xmath5 breaking effects by the quark masses to first order , and the last two terms provide the @xmath5 breaking by the electromagnetic interaction , the first one representing the electromagnetic self - energy of the light antiquark , and the second one is the electromagnetic interaction between light antiquark and heavy quark . throughout
, @xmath17 represents the heavy quark pole mass .
up to @xmath33 , the @xmath16 expansion gives for @xmath34 : @xmath35 for dimensional purposes , we use the @xmath36-meson mass as the reference mass scale .
the explicitly displayed terms @xmath37 and higher represent the mass difference produced by the strong hyperfine interaction .
@xmath38 has a non - trivial dependence on @xmath17 that results from the qcd running of the effective heavy quark operator in the @xmath16 expansion associated with it .
the heavy quark effective lagrangian in the @xmath16 expansion is given in standard notation by @xcite : @xmath39 where the third term , which couples to the heavy quark spin , gives rise to @xmath40 .
the coefficient @xmath41 , where @xmath42 is the renormalization scale , is given by @xcite : @xmath43 with the initial condition given by the matching to full qcd @xcite : @xmath44 here @xmath45 , @xmath46 , and @xmath47 is the first coefficient of the @xmath48-function with @xmath49 the number of flavors lighter than @xmath17 . because @xmath40 is proportional to @xmath50
, we can express it in the following renormalization group invariant form : @xmath51 where @xmath52 is @xmath17 and @xmath42 independent . in what follows ,
all over - lined coefficients are @xmath17 and @xmath42 independent .
the term proportional to @xmath53 receives contributions from two terms in the heavy quark lagrangian at @xmath33 @xcite .
taking into account the running of @xmath53 with @xmath17 is thus impossible in this analysis .
fortunately , this is not important because the @xmath54 terms play a minor role in the @xmath2 mesons , and thus neglecting the running of the @xmath53 is a good approximation . for the light quark mass effects , @xmath55
is defined in @xmath56 scheme and one has : @xmath57 because the spin - independent term in the @xmath58 heavy quark lagrangian is scale independent , @xmath59 and @xmath60 are independent of @xmath17 and their dependence on @xmath42 is given by the running of @xmath55 , _
i.e. _ @xmath61 , @xmath62 . on the other hand , because @xmath63 is proportional to @xmath50 , it has an extra running factor similar to that of @xmath40 : @xmath64 the @xmath42 dependence given above is rather immaterial in our analysis , where the same @xmath42 is used for @xmath0 and @xmath2 mesons . in the case of electromegnetic effects , we need to discuss seperately the two terms . the coefficients of the self - energy can be expressed in the following most general form : @xmath65 we note that @xmath66 will run with @xmath17 in a similar form as @xmath40 .
the spin of the heavy quark will affect very little the self - energy term , and therefore @xmath66 will be small . as shown later ,
@xmath66 can be eliminated because of linear dependencies , which means that the effect associated with it can not be distinguished from other effects on the meson masses . on the other hand , the electromagnetic terms involving the interaction between the light degrees of freedom and the heavy quark
have the form : @xmath67 @xmath68 and @xmath69 give the leading and subleading in @xmath16 contributions to the coulomb interaction , and both are scale independent as it is known from the renormalization of current operators of the heavy quark , such as the electromagnetic current .
the electromagnetic hyperfine effect proportional to @xmath70 can be expressed as follows : @xmath71 it receives contributions from two general types of diagrams shown in fig ( 1 ) . the first term results from the coupling of the photon to the heavy quark spin and therefore it has no @xmath17 dependence , while the second term corresponds to the coupling of a gluon to the heavy quark spin , and is therefore proportional to @xmath50 and thus @xmath17 dependent .
in this analysis , we consider the five different mass splittings possible in each multiplet , _ i.e. _ @xmath72 , @xmath73 , @xmath74 , @xmath75 , @xmath76 and similarly for b - system .
the mass formuli leave one parameter independent mass relation , which reads : @xmath77 where we denote @xmath78 .
this mass relation is violated by terms @xmath79 , @xmath80 , @xmath81 , and @xmath82 .
note that this relation was discovered in ref.@xcite , where the evolution factor @xmath83 was not included .
if one disregards the term @xmath33 in the strong hyperfine interaction , _ i.e. _ the term proportional to @xmath53 , one obtains an additional relation : @xmath84 the deviations from this relation are a measure of the importance of the @xmath85 term in the hyperfine interaction . in the mass formuli
, there is a total of twelve parameters that enter in mass differences .
since there are ten mass differences and one parameter free mass relation , there must be three linearly dependent terms in the mass formuli that we can eliminate .
the linear dependencies are such that @xmath86 and @xmath69 can be absorbed into @xmath87 and @xmath68 , and @xmath88 into @xmath38 and @xmath89 .
since no dependencies appear if one stays at leading order in @xmath16 , it is natural to eliminate sub - leading terms .
we , therefore , eliminate therefore @xmath86 , @xmath90 and @xmath69 .
the linear dependencies imply that , in this analysis , one can not determine the @xmath16 corrections to the self - energy and to the coulomb effects independently from other effects . in what follows , we therefore set : @xmath91 the quark masses are the key input parameters in the mass formuli .
if we would only keep up to @xmath58 terms , the ratio @xmath92 would determine @xmath93 for @xmath94 gev .
there are more accurate determinations of the ratio @xmath4 from the analysis of charmonium and bottomonium @xcite that give @xmath95 and @xmath96 gev . with this ratio ,
it becomes necessary to include the @xmath33 term in eqn .
note that eqn .
( 12 ) could be used to extract the ratio @xmath4 , but it requires a precision in the mass differences involved that is far beyond the current precision . for the light quark mass inputs ,
we only need their ratios , namely , @xmath97 where @xmath98 , and @xmath99 , which have been extensively studied in chiral perturbation theory @xcite . the first ratio is obtained from the ratio @xmath100 , which after next to leading order chiral corrections , gives a value of @xmath101 , and the second ratio requires the input of isospin breaking observables , in particular the mass ratio @xmath102 , and the rates for @xmath103 , with a result @xmath104 @xcite . for further reference
, this latter ratio corresponds to having the electromagnetic mass difference @xmath105 mev .
the final input is @xmath106 required by @xmath107 ; we use @xmath108 mev . with our inputs for the heavy quark masses ,
we obtain for the left hand side of eqn .
( 13 ) a value equal to @xmath109 , which gives some evidence for the need of the @xmath33 term in eqn .
( 2 ) . for eqn . ( 13 ) to hold , it would be necessary to have @xmath93 . from eqn .
( 12 ) we obtain the combination : @xmath110 which is not known experimentally because the mass difference @xmath15 has not been established separately from the one for charged ones .
the improvement over similar prediction given in @xcite is primarily due to the improved accuracy of the various inputs , especially the heavy quark masses and the running effect characterized by the factor @xmath83 . in table
i , we give the results of our fits , displaying the partial contributions , and in table ii , the effects are combined . for the strong hyperfine contributions , we see that the @xmath33 effect tends to reduce the contribution from the leading @xmath58 term by up to 20% in d - mesons and by up to 8% in b - mesons .
the errors quoted for the individual terms are rather large which indicate a strong correlation .
this is displayed in fig.(2 ) .
the error due to the uncertainty in @xmath4 manifests itself in the individual strong hyperfine terms where it is approximately @xmath111mev in the d mesons .
this indicates that , at present , the @xmath33 strong hyperfine effects can not be established with good precision . clearly , the combined hyperfine effects are very precise as shown in table ii . the leading order contributions by the light quark masses , which are independent of the meson s spin , _
i.e. _ the contributions proportional to @xmath59 , are well determined by the fit , the error being dominated by the errors in the input light quark mass ratios .
the @xmath58 corrections proportional to @xmath60 can also be determined quite easily : because the spin dependent term @xmath112 turns out to be rather small , the @xmath60 term is almost entirely determined in terms of the combination @xmath113 . in particular , its sign is positive because @xmath114 .
this is a puzzling fact : if one tries to interpret the @xmath60 term from the point of view of a non - relativistic constituent quark model , the dependence of @xmath115 on the light quark masses will be through the reduced mass of the @xmath116 system , and because the @xmath117 correction to the reduced mass is negative , the sign of the @xmath60 term should be opposite to that of the @xmath59 term .
this is indication of a clear departure from the non - relativistic quark model picture which we have not seen addressed in the literature . to complete the light quark mass effects ,
the terms proportional to @xmath112 provide the @xmath5 breaking effects to the strong hyperfine interaction , and are determined with an error of about 10% , which is remarkable . for instance , in the case of the @xmath76 mass difference , it gives an upward shift of @xmath118 mev .
the combined effects of the light quark masses , displayed in table ii , show that they are determined by this analysis with an accuracy that is in general better than 10% .
there is , however , one important and still unresolved problem concerning the light quark mass effects , and this has to do with the non - analytic contributions proportional to @xmath81 @xcite , which are expected to be large @xcite , according to the estimated value of the coupling @xmath119 @xcite that gives the amplitudes @xmath120 .
a consistent analysis should include up to the @xmath121 effects , which is beyond the current analysis .
this problem , therefore , introduces some uncertainty in the determination of the light quark mass effects that is difficult to estimate . as mentioned earlier , we can separate the electromagnetic effects into self - energy and coulomb plus hyperfine type terms .
the self - energy has only a spin independent piece and it represents an effect of less than 1 mev , and is determined with about 30% error .
it has the same sign and comparable magnitude to results from calculations based on the cottingham formula for electromagnetic mass shifts in a vmd approximation @xcite .
the effect of the coulomb interaction is given by the term proportional to @xmath68 and its subleading piece proportional to @xmath69 , as explained earlier , has been absorbed into other terms .
the fit determines the coulomb effect with an error of 11% .
the electromagnetic hyperfine effects are associated with the two parameters @xmath122 and @xmath123 . the input for
@xmath124 does not differentiate between the neutral and charged mesons ; if it is identified with the neutral mesons , then one of these two parameters can be eliminated as a consequence of the relation eqn .
( 12 ) . for this reason ,
in the fits , we carried out identifying @xmath125 ; one finds a large correlation between these two parameters , which requires that we keep only one of them .
we keep @xmath122 , which amounts to ignoring the @xmath17 dependence in @xmath123 , absorbing the rest of it into @xmath122 .
our analysis is , therefore , insensitive to the qcd running of the electromagnetic hyperfine effects , which is not surprising .
we have moreover checked that our results are almost insensitive to the interpretation of the input @xmath124 as an arbitrary combination of the charged and neutral mass differences .
the hyperfine effects are significant in the d - mesons , for instance , in the @xmath72 case , it is about 60% of the coulomb effect . on the other hand , in the b - system , the hyperfine effects are much smaller than the experimental uncertainties in the mass differences .
comparison , with the calculations in ref.@xcite , shows agreement with the results in the elastic approximation to the cottingham formula using vmd .
note that the inelastic contributions in the cottingham formula that correspond to the interaction coulomb and hyperfine terms are suppressed by @xmath16 and , therefore , to the order we are working here , they can be neglected .
it is instructive to make some comparisons .
the electromagnetic shifts for the pseudoscalar mesons are @xmath126 mev , and @xmath127 mev , which are similar within errors to @xmath128 mev .
this is however a bit of a coincidence , as it can be seen from the results obtained for the different contributions , which are in the coulomb and hyperfine cases very different in the @xmath0 and @xmath2 mesons .
if we consider the vector mesons , we obtain @xmath129 mev , and @xmath130 mev , which we can compare with @xmath131 mev ( note that , as in the case of pions , the mass difference between charged and neutral @xmath36 meson is purely electromagnetic ) .
one important observation is that the uncertainties in the ratios of light quark masses only have a noticeable effect in the self - energy terms , while the uncertainty in @xmath4 leaves the results for the electromagnetic effects virtually unchanged .
the isospin mass splittings due to @xmath132 and @xmath133 are : @xmath134 mev , @xmath135 mev , @xmath136 mev , and @xmath137 mev .
the difference between the last two is the negligible isospin breaking induced on the strong hyperfine interaction in the @xmath2 mesons .
this effect is larger and significant in the @xmath0 mesons as shown by the difference between the first two mass differences .
finally , one can give an accurate prediction : @xmath138 mev .
the combination , @xmath139 mev , gives the difference between @xmath0 and @xmath2 mesons of the @xmath58 spin independent part of the isospin breaking by the quark masses .
the electromagnetic part of the same combination is substantially larger and equal to @xmath140 mev .
a similar analysis is straightforward for the mass splittings due to @xmath141 . in summary
, we have analyzed the different contributions to the mass splittings in heavy ground state mesons .
the analysis shows that , with the current empirical accuracy of the heavy meson masses , one can determine these contributions with significant precision even at the sub - leading order in @xmath16 .
the results obtained here can be useful for constraining models of heavy mesons , and perhaps also for lattice qcd calculations of heavy mesons masses where it is possible to study the light quark mass dependence .
we would like to thank n. brambilla , j. soto and a. vairo for useful discussions on heavy quark masses , and j. soto for comments on the manuscript .
this work was supported by the department of energy through contract de - ac05 - 84er40150 ( jlg ) , and by the national science foundation through grant phy-0300185 and phy-0555559 .
99 j. l. rosner and m. b. wise , phys . rev .
* d47 * , ( 1993 ) 343 .
particle data group , w .- m .
yao et al . , j. phys .
* g33 * , ( 2006 ) 1 .
e. eichten and b. hill , phys . lett . *
b243 * ( 1990 ) 427 .
+ g. amors , m. beneke and m. neubert , phys . lett .
* b401 * , ( 1997 ) 81 .
balzereit and th .
ohl , phys . lett .
* b386 * , ( 1996 ) 335 .
+ m. finkemeier and m. mcirvin , phys
* d55 * , ( 1997 ) 377 .
+ m. finkemeier , h. georgi and m. mcirvin , phys . rev . *
d55 * , ( 1997 ) 6933 .
w. bauer and a. v. manohar , phys . rev .
* d57 * , ( 1998 ) 337 .
a. pineda and f. j. yndurain , phys . rev .
* d58 * , ( 1998 ) 094022 .
+ m. eidemller , phys . rev . *
d67 * , ( 2003 ) 113002 .
+ n. brambilla , a. pineda , j. soto and a. vairo , rev .
* 77 * ( 2005 ) 1423 , and references therein .
h. leutwyler , phys . lett . * b378 * , ( 1996 ) 313 .
j. l. goity , phys .
rev . * d46 * , ( 1993 ) 3929 .
l. randall and e. sather , phys .
b303 * , ( 1993 ) 345 j. l. goity and w. roberts , phys . rev
. * d60 * , ( 1999 ) 034001 .
j. l. goity and w - s .
hou , phys . lett .
* b282 * , ( 1992 ) 243 . j. l. goity , phys . lett . *
b303 * , ( 1993 ) 337 . .
the mass shifts due to the different terms are labeled by the corresponding coefficient in the mass formuli and are in the in units of mev .
the errors include the uncertainties in the input quark mass ratios . in the fit @xmath124
is interpreted as @xmath142 . [
cols="^,^,^,^,^,^,^,^,^",options="header " , ] | the contributions to heavy meson mass differences by the strong hyperfine interaction , the light quark masses and the electromagnetic interaction are obtained from the empirical values of the @xmath0 , @xmath1 , @xmath2 and @xmath3 masses by means of a mass formula based on the heavy quark mass expansion .
the three different types of contributions are determined with significant accuracy to next to leading order in that expansion . |
magnetohydrodynamic ( mhd ) mechanisms are often invoked to explain the launching , acceleration and collimation of jets from young stellar objects , x - ray binaries , active galactic nuclei ( agns ) , microquasars , and quasars ( see , _ e.g. _ , * ? ? ?
* and references therein ) .
strongly magnetized jets , particularly those with a strong toroidal field encircling the collimated flow , are often referred to as `` current - carrying '' or `` poynting flux - dominated '' ( pfd ) jets .
a large current flowing parallel to the jet flow is responsible for generating a strong , tightly wound helical magnetic field .
the global picture of a current - carrying jet with a closed current system linking the magnetosphere of the central engine and the hot spots was introduced by @xcite and applied to agn double radio sources .
this closed current system includes a pair of current circuits , each containing both a forward electric current path ( the jet flow itself , with its toroidal magnetic field , toward the lobe ) , and a return electric current path ( along some path back to the agn core ) .
recent high resolution observations , in x - ray , optical and radio bands , show that many agn and quasar jets display wiggles / kinks , which may be associated with a helical motion of the underlying bulk flow , in wide spatial ranges from sub - parsec to kpc scales ( _ e.g. _ , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
many ideas have been proposed for these helical distortions : mhd instabilities such as the kelvin - helmholtz instability ( khi ) , the current - driven instability ( cdi ) , and the pressure - driven instability ( pdi ) , or the precession of the jet ejection axis due to the existence of a binary black hole @xcite , or an encounter with another galactic core .
it is well known that a cylindrical plasma column with a helical magnetic configuration is subject to the @xmath1 ( sausage ) mode of pdi , the @xmath0 ( kink ) mode of cdi , and the other higher order modes , where @xmath2 is the azimuthal ( @xmath3 ) mode number ( see , _
e.g. _ , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for details ) .
cdi is driven by parallel ( axial ) current flows and can exist even in a zero pressure , force - free plasma . on the other hand ,
pdi is powered by perpendicular current flows and it is often known as the interchange ( internal ) modes .
khi is considered to be very important at the shearing boundary between the flowing jet and the external medium , particularly in the kinetic flux - dominated ( kfd ) jets . in principle ,
cdi and khi have both external ( surface ) and internal ( body ) modes .
pfd jets should be especially susceptible to cdi rather than pdi and khi .
this is because the strong axial electric current is present inside the pfd jets , which is responsible for the highly wound helical field configurations , and the lorentz force plays an important role in the jet dynamics .
the investigation of the destructive influence of cdi on pfd jet flows is an important avenue of research in the study of astrophysical jets .
much more attention on studies of stabilities ( disruption of the interior / exterior of astrophysical jets ) have given to khi than to cdi .
this is because jets were believed to be super - alfvnic or super - fast magnetosonic kfd flows ( _ i.e. _ , kinetic energy was expected to greatly exceed the magnetic energy ) .
however , it is not necessarily always the case ; a super - alfvnic or super - fast magnetosonic pfd flow can exist in both nonrelativistic @xcite and relativistic @xcite mhd outflow solutions .
analytic studies on the cdi as a possible explanation for jet disruption have been performed as well @xcite .
linear analyses of nonrelativistic force - free jets , which are thermally confined by the external nonmagnetized medium , were performed in super - fast @xcite or trans - fast @xcite magnetosonic regimes .
they found that the growth rate of cdi is always substantially smaller that those of khi in their specific treatment .
only for quite small fast magnetosonic numbers , @xmath4 , both become comparable .
linear analyses of the relativistic force - free jets , in case of constant - longitudinal field @xmath5 along the jet axis , were performed by @xcite .
they concluded that such a jet is stable against cdi , while @xcite analyzed the relativistic force - free jets with @xmath5 decreasing outwards and found that they are unstable .
nonlinear development of the cdi for cold super - fast magnetosonic jets has been studied by @xcite , based on the linear analysis @xcite .
it was found that the current density is redistributed within the inner part of the jet radius on a characteristic time scale of the order of the alfvn crossing time in the jet frame .
nothing in their numerical results indicated a possible disruption of the jet by the cdi sausage ( @xmath1 ) or kink ( @xmath0 ) modes ( see also * ? ? ? * ; * ? ? ? * ) . in general , these linear / nonlinear considerations conclude that `` current - driven instabilities are unlikely to be dangerous for the integrity of astrophysical jets '' @xcite and
this has been widely accepted in theoretical / observational communities . in the nonlinear regime
, an interplay between khi and cdi can occur in the helically magnetized flow even with slightly super - fast magnetosonic speed @xmath6 @xcite .
this nonlinear interaction can contribute to jet survival , and the large - scale magnetic deformations associated with the cdi mode development can effectively saturate khi surface vortices and prevent jet disruption . by contrast , the first attempt to investigate the nonlinear behavior of visible distortion in kfd jets along the large - scale magnetic field by the cdi kink mode was performed by @xcite to model herbig haro objects . in a similar study , @xcite examined the formation of wiggled structures by the @xmath0 mode in pfd jets to apply for agn jets .
their numerical result is applicable for the systematic distribution of the observed rotation measure in 3c 449 @xcite , associated with the jet deformation @xcite .
the rotating pfd / kfd jets can be potentially stabilized against the @xmath0 mode beyond the original kruskal - shafranov criterion , however , sudden destabilization of rotating jets due to the @xmath0 mode can occur via the angular momentum loss triggered by mhd shock waves @xcite .
the purpose of the present paper is to discuss the stability properties of `` magnetic tower '' jets , which have not been studied in the past work .
of particular interest here is the nonlinear growth of the cdi kink mode .
the cdi kink mode in magnetic tower jets may be distinguished from that in the classical mhd jets along the large - scale magnetic field @xcite , while it should be comparable directly to `` non - disrupting '' models @xcite , in which the mhd jets are assumed to be in pressure equilibrium with an unmagnetized ambient medium .
this is the third of our series of papers to examine the nonlinear magnetic tower jets .
the first of our series , ( * ? ? ?
* hereafter paper i ) , described the basic assumptions and the approaches in the numerical modeling of magnetic tower jets .
the evolution of the tower jets in a constant density / pressure background has been examined there .
the second of our series , ( * ? ? ?
* hereafter paper ii ) investigated the global structure of magnetic tower jets in the gravitationally stratified atmosphere ( a more realistic astrophysical situation ) , in terms of the mhd waves structure , the radial force equilibrium , and the collimation .
in the present paper , the third of our series , we investigate the stability properties of magnetic tower jets .
we examine our numerical results in two cases ; one without initial perturbations and another with initial perturbations .
the former exhibits the quasi - axisymmetric evolution without any growth of mhd instabilities , while the latter shows non - axisymmetric evolution by the cdi modes .
this paper is organized as follows . in 2 , we outline our numerical methods . in 3 , we describe our numerical results .
discussion and conclusion are given in 4 and 5 .
we solve the nonlinear system of time - dependent ideal mhd equations numerically in a 3-d cartesian coordinate system @xmath7 .
the basic numerical treatments ( including the mhd numerical scheme ) is essentially the same as that in paper i and ii . we assume an initial hydrostatic equilibrium in the gravitationally stratified medium , adopting an iso - thermal king model @xcite .
the magnetic flux and the mass are steadily injected in a central small volume during a certain time period .
since the injected magnetic fields are not force - free , they will evolve as a `` magnetic tower '' and interact with the ambient medium . in this paper
, we present mainly two different cases : one is called the `` unperturbed case '' in the following discussion , which is a run without any initial perturbation to the background initial profiles .
this simulation is taken from the run presented in paper ii .
the other is called the `` perturbed case '' , where a finite amplitude perturbation ( a few percent of the background sound speed ) is given to the velocities of the background gas .
the total computational domain is taken to be @xmath8 .
the numbers of grid points in the simulations reported here are @xmath9 in the unperturbed case and @xmath10 in the perturbed case , where the grid points are assigned uniformly in the @xmath11 , @xmath12 , and @xmath13 directions .
we normalize physical quantities with the unit length scale @xmath14 , the unit density @xmath15 , the sound speed @xmath16 as the typical velocity field in the system , and other quantities derived from their combinations , _
e.g. _ , the typical time @xmath17 as @xmath18 , etc .
normalizing factors are @xmath19 kpc , @xmath20 cm / s , @xmath21 yrs , @xmath22 g/@xmath23 , the unit pressure @xmath24 as @xmath25 dyn/@xmath26 , and the unit magnetic field @xmath27 as @xmath28 @xmath29 g .
we take @xmath30 and @xmath31 as the initial quantities at the origin @xmath32 and the @xmath33 and @xmath34 at the origin in the simulations are set to unity .
the initial sound speed in the simulation is constant , @xmath35 , throughout the computational domain .
the simulation domain is from @xmath36 to @xmath37 kpc , with @xmath38 for the unperturbed case and @xmath39 for the perturbed case , which correspond to @xmath40 and @xmath41 respectively .
an important time scale in the system is the sound crossing time @xmath42 , corresponding to a typical time scale @xmath43 myrs .
therefore , @xmath44 is equivalent to the unit time scale @xmath45 myrs . in the king model
we use here , we adopt the cluster core radius @xmath46 to be @xmath47 ( i.e. , 20 kpc ) and the slope @xmath48 to be @xmath49 in the unperturbed case and @xmath50 in the perturbed case .
the injections of magnetic flux , mass and its associated energies are the same as those described in paper i. the ratio between the toroidal to poloidal fluxes of the injected fields is characterized by a parameter @xmath51 , which corresponds to @xmath52 times more toroidal flux than poloidal flux .
the magnetic field injection rate is described by @xmath53 and is set to be @xmath54 .
the mass is injected at a rate of @xmath55 over a central volume with a characteristic radius @xmath56 .
magnetic fluxes and mass are continuously injected for @xmath57 , after which the injection is turned off .
these parameters correspond to a magnetic energy injection rate of @xmath58 ergs / s , a mass injection rate of @xmath59/yr , and an injection time @xmath60 myrs .
we use the outflow boundary conditions at all outer boundaries . note that for most of the simulation duration , the waves and magnetic fields stay within the simulation box , and all magnetic fields are self - sustained by their internal currents . as examined in paper ii
, the magnetic tower jets can propagate without any visible distortion throughout the time evolution ( see fig . 1 of paper ii ) .
the magnetic tower has a well - ordered helical magnetic configuration ; a tightly wound central helix is going up inside the tower jet and a loosely wound helix is coming back at the outer edge of magnetic tower jet ( see fig .
3 of paper ii ) .
the axial current flow @xmath61 displays a closed circulating current system in which one path flows along the central axis ( the `` forward '' current ) and another conically shaped path that flows outside ( the `` return '' current ) ( see fig . 4 of paper ii ) . here
, we define the forward current density @xmath62 and the return current density @xmath63 as follows ; @xmath64 note that @xmath62 ( and @xmath65 ) is a vector and can have additional radial and azimuthal components @xmath66 and @xmath67 ( @xmath68 and @xmath69 ) with any sign . according to the well - known kruskal - shafranov ( k - s ) criterion @xcite , a magnetized flux tube ( along the @xmath13 axis ) is stable to the kink ( @xmath0 ) mode as long as the magnetic twist angle @xmath70 is below some critical value @xmath71 , @xmath72 where
, @xmath73 is the length of the current - carrying magnetic flux system , @xmath74 is the cylindrical radius of the flux tube , and @xmath5 and @xmath75 are the axial and azimuthal field components in a cylindrical coordinate , respectively . in the original k - s criterion , @xmath76 is equal to @xmath77 . the effect of `` line - tying '' , i.e. , fixing the radial motion of the magnetic fluxes at the boundary , however , could raise the stability threshold .
( in the solar coronal loops , the foot points of magnetic fluxes can be anchored in a high density region as chromosphere . )
the modified critical values for a force - free magnetic loop configuration in solar coronal loops are between @xmath77 and @xmath78 ( e.g. , * ? ? ?
* ; * ? ? ?
for the configuration of an astrophysical jet , which has @xmath79 , inequality ( [ eq : k - s_limit ] ) can be easily violated .
if the mhd jets originate from an accretion disk , the bottom part of jets may be line - tied in the disk and the top part of jets can be free against the radial motion ; i.e. , a `` semi '' line - tying configuration .
( further argument of the stabilizing effect by semi line - tying will be presented in 4.2 . ) for lack of detailed linear analysis of `` semi '' line - tying magnetic configurations , we still define @xmath80 , which should only be regarded as a guide for gauging the stability , rather than a precise threshold . here , we replace @xmath73 by a spectrum of axial wavelengths @xmath81 , and define the critical wavelength to be the one that is marginally stable : @xmath82 the k - s stability criterion ( eq . [ [ eq : k - s_limit ] ] ) then becomes @xmath83 so that wavelengths greater than @xmath84 could be unstable to the cdi kink mode . to better diagnose the non - axisymmetric distortion , we perform the power spectrum analysis of the magnitude of the current density @xmath85 . we analyze the modal structure of @xmath85 by performing the volume - averaged fourier transform : @xmath86 where @xmath87 and @xmath88 is the cylindrical volume that encloses @xmath89 @xmath90 we use the values @xmath91 , @xmath92 , @xmath93 , and @xmath94 in the `` upper '' volume and use @xmath95 and @xmath96 in the `` lower '' volume .
the quantity @xmath97 is a function of the azimuthal mode number @xmath2 and axial wave number @xmath98 ( corresponding to a characteristic wave length @xmath81 ) , and it is a function of time as well . finally , we identify the power spectrum as @xmath99 \right\}^{2 } + \left\{{\rm i m } \bigl[f(m,\,k ) \bigr ] \right\}^{2},\end{aligned}\ ] ] @xmath100 \nonumber \\ & & \quad = \frac{1}{v_{\rm cl } } \int\int\int_{v_{\rm cl } } \label{eq : fourier_5 } & & { \rm i m } \bigl[f(m,\,k ) \bigr ] \nonumber \\ & & \quad = \frac{1}{v_{\rm cl } } \int\int\int_{v_{\rm cl } } examine the time - dependent behavior of the power spectrum and the growth of each mode .
we first re - examine the numerical result of paper ii as an unperturbed case , from a viewpoint of stability .
for that simulation , we did not impose any initial perturbations to the background state .
so the only available perturbations are from the numerical noise in the double precision computations .
figure [ fig : t7.5_jz_wlc ] shows the snapshots of @xmath101 ( _ top _ ) and @xmath102 ( _ bottom _ ) in the final stage ( @xmath103 ) .
again , we can identify the circulating path of axial current flow @xmath61 which consists of both the forward axial current density @xmath104 and the return current @xmath105 in the _ top _ panel ( @xmath106 lies around the central axis , while @xmath107 surrounds @xmath108 ) .
the maximum strength of @xmath109 is much larger than that of @xmath110 , indicating that a tightly wound central helix is going up at the center of the tower and a loosely wound helix is coming back at the tower edge .
the paths of forward and return currents are approaching one another around the mid - plane @xmath111 . on the other hand , the path of return current is radially expanding at the upper axial position @xmath112 .
this causes a _ separation _ of current flow paths in a thermally confined mhd jet with a helical magnetic field .
( this plays an important role in the excitation of the external kink mode and it will be discussed in 4.3 . ) _ top _ : distribution of the axial current density @xmath61 in the @xmath113 plane at @xmath103 in the unperturbed case ( paper ii ) . _
bottom _ : distribution of the critical wavelengths @xmath102 at @xmath103 in the @xmath113 plane , corresponding to the original kruskal - shafranov criterion for cdi @xmath0 mode in the unperturbed case ( paper ii ) . ]
as seen in the _ bottom _ panel , @xmath84 has a range 2.0 @xmath114 6.0 in the core part of @xmath115 , and has no axial dependency except at the tower front . while @xmath84 has no range in any part of @xmath116 , meaning that the cdi kink mode is unlikely to grow at the return current region .
taking the whole part in the axial direction into consideration , the maximum length of current - carrying column is @xmath117 at @xmath103 , which is larger than the distributed @xmath84 in the core part of @xmath108 .
however , the non - axisymmetric displacement does not seem to occur as seen in both panels .
we next show the result of the power spectrum analysis of @xmath118 in order to confirm the non - existence of the growing kink mode .
figure [ fig : mode_power_p19 ] shows the space - time @xmath119 diagrams of the power spectrum @xmath120 for @xmath0 in the unperturbed case .
we can not detect any growing non - axisymmetric mode ( @xmath0 ) over the wider range of axial wave length ( @xmath121 ) .
the time variation of the power spectrum of @xmath122 with a selected wavelength @xmath123 in fig .
[ fig : growth_rate_p19 ] . throughout the time evolution ,
the power in @xmath0 mode is always below the axisymmetric ( @xmath1 ) one .
note that the variation of the @xmath1 mode ( seen in both volumes around @xmath124 ) is due to the axisymmetric deformation of the magnetic tower into a slender - shaped structure by the gravitational contraction radially .
the time variation of the azimuthal fourier power ( a natural log scale ) in modes @xmath125 with a specified wave length @xmath123 in the unperturbed case ( paper ii ) . _
top _ : for the upper volume @xmath126 . _ bottom _ : for the lower volume @xmath127 . ]
it may seem puzzling that , for the unperturbed case , the magnetic tower jet is stable even with @xmath128 . here , we consider a possible mechanism of the stabilizing effect in our results .
figure [ fig : wlcrit_evo ] shows the time evolution of the k - s critical wavelength @xmath84 in the unperturbed case . a gradual increase of the distributed @xmath129 in the forward axial current - carrying column @xmath106 can be observed .
this corresponds directly to a decrease of the ratio @xmath130 . because the injection of magnetic fluxes into the central volume in the system has been turned off at @xmath131 , the twisted magnetic fluxes are naturally relaxed during the expansion of the magnetic tower .
the threshold length of the marginally stable of @xmath106 in the axial ( @xmath13 ) direction becomes longer as the magnetic tower proceeds .
therefore this `` dynamical relaxation '' of magnetic twists plays an important role in the stabilization in our current results .
we will discuss other effects on the stabilization in 4.2 .
time evolution of the k - s critical wavelength @xmath84 in the @xmath113 plane at @xmath132 ( _ top _ ) , 7.5 ( _ middle _ ) , and 10.0 ( _ bottom _ ) in the unperturbed case ( paper ii ) . ]
we next examine the non - axisymmetric evolution of the magnetic tower jet in a gravitationally stratified atmosphere . to excite non - axisymmetric modes ( @xmath133 ) , a small random velocity perturbation ( 4% of initial sound speed ) has been added into the system as an initial condition .
figure [ fig : de_final ] shows a snapshot of the density @xmath33 at the final stage @xmath134 ( see fig
. 1 of paper ii for comparison ) in the perturbed case . even though the configuration of the magnetic tower is non - axisymmetric , the main features remain similar to the unperturbed case . in general , the structure of the magnetic tower consists of both a well - collimated `` body '' and radially extended `` lobes '' .
several key features in the magnetic tower jets ( the mhd wave structures , the heating process at the tower front , the cylindrical radial force equilibrium at the tower edge , and dynamically collimating process ) are the same as the unperturbed case ( see paper ii for details ) .
the alfvn speed becomes large ( about three times of local sound speed ) , while the plasma @xmath135 becomes small ( @xmath136 ) inside the density cavities due to the expansion of magnetic fluxes .
the bulk flow inside the magnetic tower is sub - sonic and sub - alfvnic , while the fronts of the preceding hydrodynamic shock and the magnetic tower can propagate super - sonically .
similar to fig .
[ fig : mode_power_p19 ] , we present the space - time @xmath119 diagrams of the power spectrum @xmath137 in the perturbed case for the @xmath0 and 2 in fig .
[ fig : mode_power_b3 ] .
clearly , the growth of the @xmath0 mode can be seen over a wide range of @xmath81 after @xmath138 in both upper and lower volumes .
furthermore , the @xmath139 ( elliptical ) mode also grows at a relatively short wavelength around @xmath140 after @xmath141 . to identify the growth rates of different modes ,
we show the time variation of the power spectrum of @xmath122 in the perturbed case with a selected wavelength @xmath123 in fig .
[ fig : growth_rate_b3 ] ( similar to fig .
[ fig : growth_rate_p19 ] ) .
the @xmath0 mode in both volumes exhibits an approximate exponential growth on a dynamical timescale during @xmath142 . at later stages , after the @xmath0 mode grows above the axisymmetric @xmath1 mode , its growth rate decreases slightly , but still keeps growing exponentially to saturation at the final stage .
the @xmath139 mode also shows the similar behavior with the @xmath0 mode at the earlier stage around @xmath143 , but it saturates at or below the level of the @xmath1 mode .
similar to fig .
[ fig : growth_rate_p19 ] , but for modes @xmath144 in the perturbed case . straight lines are derived by fitting the slopes of the @xmath0 mode growth to a linear function . _
top _ : for the upper volume @xmath126 .
_ bottom _ : for the lower volume @xmath127 . ]
a linear fit using a least - square method has been performed on the initially exponential growth for the @xmath0 mode to extract a linear growth rate @xmath145 the estimated growth rates in the perturbed case are also given in fig .
[ fig : growth_rate_b3 ] .
the growth rate in the upper volume appears to be a bit stronger than that in the lower volume .
we confirm the distributed alfvn speed @xmath146 in the surfaces of @xmath118 is of the order of @xmath147 in our normalized unit during the time evolution .
the spatial length scale @xmath148 of the non - axisymmetric ( @xmath149 ) distortions in the radial direction is of the order of @xmath150 , as seen in fig .
[ fig : de_final ] . therefore , the inverse of the alfvn crossing time @xmath151 in the distorted part of magnetic tower jet is given by @xmath152 this is , in general , consistent with the timescales of the growing @xmath0 mode derived from fig .
[ fig : growth_rate_b3 ] .
so , our results lead us to conclude that the non - axisymmetric distortion in the magnetic tower jet is caused dominantly by the normal current - driven kink ( @xmath0 ) mode . snapshots of the axial current density @xmath61 in the @xmath113 plane at @xmath153 ( _ top _ ) and 9.5 ( _ bottom _ ) in the perturbed case . two types of growing kink modes ( `` external''/internal ) are illustrated on the _ top _ panel . ]
two types of cdi modes are identified in the context of magnetically controlled fusion plasma ; the `` external '' and the `` internal '' modes @xcite .
the external mode can be identified with a radial distortion , while in the internal mode , there is no radial perturbation at the outer edge of the current - carrying column .
one of our findings in the perturbed magnetic tower jet is that the cdi external and internal modes `` co - exist '' during the dynamical evolution .
figure [ fig : jz_evo ] shows snapshots of @xmath61 in the perturbed case at @xmath153 and 9.5 . as seen in the _ top _ panel ( @xmath153 ) , the edge of @xmath154 is clearly shifted to an offset axial direction at higher latitude ( @xmath155 ) . on the other hand ,
that edge never shifts non - axisymmetrically at lower latitude ( @xmath156 ) .
however , the non - axisymmetric distribution of @xmath118 at the inner part can be detected , indicating that the internal mode is certainly growing . at the final stage ( @xmath134 : _ bottom _ ) , we can see the well - grown wiggles inside the `` lobes '' , while the edge of the `` body '' part in the magnetic tower still remains quasi - axisymmetric except for the inner part .
taking the result of fig .
[ fig : mode_power_b3 ] into consideration , the @xmath0 kink mode is dominant in the external cdi at the lobes , while the @xmath139 elliptical mode may be dominant in the internal cdi at the body of the magnetic tower .
in addition , we illustrate a 3-d view of some selected magnetic lines of force in the perturbed case at the final stage @xmath134 in fig . [ fig:3dlines ] .
it indicates that a tightly wound central helix , which goes up along the central axis , relaxes in the lobe part where the external kink mode has taken place , and a large - scale loosely wound twist is formed there .
three - dimensional view of some selected magnetic lines at @xmath134 in the perturbed case . ]
we first discuss the possibility of the interplay between cdi and khi modes in our results .
the examinations of khi in linear / nonlinear regimes have been extensively performed by p. e. hardee and his collaborators ( see , _
e.g. _ , * ? ? ?
* and references therein ) .
in general , non - axisymmetric `` surface '' ( _ external _ ) modes ( @xmath157 ) are stable against khi for sub - alfvnic flow . however
, in super - alfvnic but trans - fast magnetosonic flow , they can be unstable . only axisymmetric surface mode ( @xmath1 ) remains unstable in sub - alfvnic flow , although with a relatively small growth rate @xcite .
a potential zone of unstable region in non - axisymmetric khi surface modes should exist in downstream of the alfvn surface .
the instability criterion for non - axisymmetric khi surface modes is as follows @xcite : @xmath158 where @xmath159 is the velocity shear and @xmath160 is the _ surface _ alfvn speed ( subscript @xmath161 corresponds to the jet itself and @xmath162 corresponds to the external medium ) . on the other hand ,
the non - axisymmetric `` body '' ( internal ) modes ( @xmath163 ) can become important and affect the jet interior in the following situations : ( 1 ) if the jet velocity exceeds the super - fast magnetosonic speed @xmath164 : @xmath165 or ( 2 ) if the flow velocity is slightly below the slow magnetosonic speed @xmath166 : @xmath167 where @xmath168 is the sound speed and @xmath169 is the alfvn speed @xcite .
figure [ fig : khistable ] shows the transverse profiles in the @xmath11 direction of several quantities at @xmath153 in the perturbed case .
two different axial positions @xmath170 and 8.0 are selected for examining the khi surface / body modes .
for the surface modes , we inspect the transverse distribution of the bulk flow speed @xmath171 , the density @xmath33 , and the magnetic field strength @xmath172 .
two distinct velocity shears ( in both positive and negative @xmath11 regions ) are identified in the @xmath11 direction on @xmath170 ( _ upper - left _ ) ; an outer part @xmath173 and an inner part @xmath174 . at the outer part , @xmath175 and @xmath176 and at the inner part , @xmath177 and @xmath178 .
therefore , the inequality @xmath179 is satisfied .
similarly , we can see many distinct velocity shears in the _ upper - right _ panel .
however , even in these shears , we confirm that the inequality @xmath180 still holds everywhere .
this indicates that the khi surface modes are completely stabilized , because there is no significant velocity shear exceeding the local surface alfvn speed in the transverse ( @xmath11 ) direction . for the body modes , we inspect the transverse distribution of the bulk flow speed @xmath171 , the alfvn speed , and the quantity @xmath181 . inside the magnetic tower ,
the plasma @xmath182 ( @xmath183 ) is always less than unity during the time evolution .
therefore , the phase speed of the mhd fast mode wave is equal ( in the local magnetic field direction ) or larger ( in the oblique direction of the local magnetic field ) than the alfvn speed , @xmath184 . as seen in both _
panels in fig .
[ fig : khistable ] , the bulk flow speed is always extremely sub - alfvnic , and furthermore , the inequality @xmath185 is also satisfied in the `` body '' part of the magnetic tower jet .
these indicate that the khi body modes are suppressed as well .
we conclude that the khi modes play no role and the normal cdi modes create the non - axisymmetric structures in our results . as examined in
[ sec : a ] , the unperturbed magnetic tower jet can be stabilized against the cdi kink mode beyond the k - s criterion .
one of the certain mechanisms of the stabilizing effect , which we can identfy in our unperturbed case , is the dynamical relaxation of magnetic twists .
however , this process may contribute the _ marginal stabilization _ under an ideal situation ( no perturbation except for the numerical noise is assumed as an initial condition ) . if the system is perturbed randomly with a finite amplitude ( a few percent of the local sound speed ) , then the non - axisymmetric ( @xmath186 ) internal / external cdi modes can grow .
the disturbances in the system are reasonable ( especially when related to possible astrophysical situations ) .
therefore , the magnetic tower jet in our treatments , will eventually be subject to cdi . here
, we give further discussions on the stabilities of magnetic tower jets . in the solar coronal magnetic loops
, it is well known that the line - tying effect at the foot points ( chromosphere ) can raise the original k - s stability threshold ( dense materials can prevent the radial motion of twisted magnetic fluxes ) .
how about the case in the astrophysical jets ?
anchored magnetic fluxes to the accretion disk and/or stellar / black hole surfaces may be considered as a semi ( one side ) line - tying boundary condition in a certain dynamical stage .
an important clue to this can be derived from laboratory experiments of magnetic towers @xcite .
their quasi `` force - free '' magnetic towers can exhibit the systematic @xmath0 distortion when the current - carrying system satisfies the original k - s limit .
it might be unlikely that such a semi line - tying effect can stabilize the jet even in the astrophysical situation . in our global modeling of magnetic tower jets
, we do not solve the evolution of the central region ( an accretion disk , etc . ) .
therefore , we can not apply it directly to the agn central region .
it is still an open issue whether a semi line - tying effect can work or not in the real astrophysical jets . as discussed in paper ii , the `` forward ''
current - carrying column @xmath154 never reaches a force - free equilibrium .
instead , it will be supported by the outward - directed , local ( core of the magnetic towers ) pressure gradient force : @xmath187 during a dynamical evolution ( see fig . 10 in paper ii ) . in terms of minimization of the mhd energy integral @xcite , a finite gas pressure in the system ( plasma @xmath188 ) can contribute to the stabilization .
however this can be effective when @xmath189 , i.e. , the gas pressure is more dominant than the magnetic pressure ( for @xmath0 kink mode , see * ? ? ?
therefore , we may not expect a strong stabilization by the pressure in the magnetic tower ( @xmath190 ) .
furthermore , the outward - directed gas pressure gradient itself ( @xmath191 ) actually destabilizes the tower through the pdi ( interchange and/or ballooning ) modes , which is driven by the perpendicular currents ( in contrast to the axial [ parallel ] currents , considered mainly in this paper ) .
we , however , could not find any growth of the pdi modes , in both the unperturbed ( paper ii ) and the perturbed cases , implying no direct contribution of the pressure gradient to the destabilization by the pdi modes ( see next section for more details ) .
also , we can not confirm any direct contributions of the internal thermal pressure effects ( @xmath192 and @xmath193 ) to the destabilization by the cdi internal / external modes .
we next consider the magnetic field configuration in magnetic tower jets . in the context of magnetic controlled fusion systems ,
the helical field configuration in the tower model can be identified as the reversed field pinch ( rfp ) profile as seen in fig .
11 in paper ii .
the axial component @xmath5 has a crucial shear ( its sign reverses from `` + '' to `` - '' ) in the radial direction ( rfp in the context of astrophysical jets , see also * ? ? ?
several stabilization effects in rfp are summarized in @xcite ; ( i ) a hollow or very flat pressure profile ( @xmath194 ) are required near the central ( @xmath13 ) axis to suppress the pdi sausage ( interchange , internal ) mode ( @xmath1 ) , but ( ii ) it can be suppressed when @xmath195 .
this may be the situation in the present paper . as already discussed in the previous section ( [ sec : ds_itp ] ) , a finite pressure gradient exists to contribute to the radial force equilibrium in the core part ( around the central axis ) and this might play a role in driving the pdi modes .
however , the plasma @xmath182 is small ( @xmath196 ) , indicating that the destabilization by the pdi modes may be suppressed due to the small plamsa @xmath182 .
also , ( iii ) the strong @xmath5 shear can prevent the destabilization by the cdi internal / external kink modes ( @xmath0 ) . in our results , this shear may provide the stabilization against the cdi external modes , while it does not affect the cdi internal modes at the lower latitude @xmath197 , as seen in figs . [
fig : mode_power_b3 ] and [ fig : jz_evo ] .
furthermore , the cdi external modes can grow at the higher latitude @xmath155 .
this may be because the strong @xmath198 shear in the radial direction will be relaxed towards the jet axis due to the decreasing external pressure ( expanding the poloidal fluxes to the radial direction ) .
the external gas itself also may affect the stabilization .
the inward - directed , thermal pressure confinement and/or the gravitational contraction of the external unmagnetized gas , which surrounds the magnetic tower jets , may provide some stabilization at least against the growth of external cdi modes , although the internal cdi and pdi modes will not be affected ( for the internal pdi sausage @xmath1 mode , see * ? ? ?
in our results , the external environment may have a similar stabilizing effect on the external cdi modes to the @xmath5 shear , as discussed in the previous section ( [ sec : ds_rfp ] ) , while the internal pdi mode may be suppressed by the low internal pressure . in the magnetic tower jets ,
an external gas plays a crucial role in the stabilization of the cdi external modes . when an external gas pressure profile does not decrease ( constant ) , which is examined in paper i ( see fig .
11 in paper i for a continuous injection case ) , the magnetic tower never expands radially and both the thermal confinement by the external gas and the strong @xmath198 shear will be kept throughout the magnetic tower `` body '' ( no formation of `` lobe '' ) .
therefore , the cdi internal modes may occur , but the cdi external modes should be surpressed .
these are consistent with the cdi `` non - disrupting '' model @xcite as well .
we finally discuss the stabilization by the rotation . in our numerical treatment , we investigate almost `` non - spinning '' magnetic tower jet . from the theoretical point of view , any magnetically driven outflows ( magneto - centrifugal and/or magnetic pressure driven ) can have an azimuthal velocity component .
they possess an angular momentum provided by the accretion disk or kerr black hole magnetosphere . in the classical model of nonrelativistic mhd winds ,
the centrifugal force plays a minor role in the radial force equilibrium far from the accretion disk @xcite .
however , this is only the case in the nonrelativistic treatment . in the relativistic mhd regime ,
the situation will be modified ; semi - analytical solutions of relativistic mhd winds with self - similarity exhibit that the relativistic inertia and electrostatic forces play an important role in the radial force equilibrium at the cylindrical flow structure , extended to pc or larger scales @xcite . braided
emission - line profiles and rotational velocity in ngc 4258 indicate a pure helical motion exists along the tightly wrapped strand even in kpc scales @xcite .
the stabilizing effect of rotation against the cdi kink mode in the spinning nonrelativistic pfd / kfd jets has been investigated by numerical simulations @xcite . because of the centrifugal effect , these spinning jets can be stabilized against the @xmath0 mode beyond the point predicted by the k - s criterion .
the linear stability analysis of the relativistically rotating force - free fields in kerr geometry @xcite has been performed by @xcite .
they also show that the field - line rotation has a stabilizing effect against the kink modes satisfying the k - s criterion . in these cases ,
the presence of an azimuthal velocity component can modify the radial force equilibrium , making the jet robust against the cdi kink mode .
examinations of the stability property in a rotating magnetic tower jet are clearly desirable .
the cdi of current - carrying jets for two types of return current distributions have been discussed : ( 1 ) the return current is assumed to flow within the jet itself ( the jets are thermally confined by the external medium ) @xcite , and ( 2 ) the return current is assumed to flow around the jet , such as in a magnetized cocoon @xcite . most of the theoretical / numerical works have been devoted to the former case , in which the forward current @xmath118 and the return current @xmath199 are concentrated at the interfaces between the jet and ism / igm , which is taken to be `` unmagnetized '' .
in this situation , the growth of external cdi modes can be effectively suppressed , while the excitations of only internal cdi modes are allowed .
there is no radial displacement at the edge of the current - carrying column by the internal cdi modes , and therefore , it may be difficult to explain the off - set axial distortion of observed jets by these internal modes . as described in [ sec : int ] , this is one of the reasons why much less attention has been paid to the cdi than to the khi for the stability problems in astrophysical jets to explain the disrupted jet exterior .
however , in the gravitationally stratified atmosphere , the effect of thermal confinement on the edge of the current - carrying column will get gradually weaker as the magnetic tower grows . as a result ,
a separation of current flowing paths between @xmath154 and @xmath200 occurs , as seen in fig .
[ fig : jz_evo ] .
this implies that the edge of @xmath154 _ becomes a free boundary _ against the cdi external modes , even when it is embedded deeply inside the external , unmagnetized gas .
furthermore , the region between the outside of @xmath154 and the inside of @xmath199 becomes nearly force - free @xmath201 .
therefore , the cdi external modes can grow in a dynamical time scale to form the systematic wiggles .
this is also true even in the spinning pfd / kfd jets under the large - scale magnetic field @xcite . on the other hand ,
even if the gas pressure in the external , unmagnetized gas is high enough to suppress the growth of the cdi external kink mode , then the cdi internal kink mode can grow to form the internal non - axisymmetric structure , as we exhibited in fig .
[ fig : jz_evo ] .
therefore , we believe that the external cdi modes are one of the possible mechanisms to explain the disruption of current - carrying mhd jets and it could be applied to the wiggling agn jets .
stability properties of magnetic tower jets have been examined by performing 3-d mhd simulations .
our numerical results show that the magnetic tower jets in the gravitationally stratified atmosphere could survive the current - driven kink instability beyond the kruskal - shafranov criterion even in the nonlinear regime .
one of the potentially stabilizing effects on a traditionally unstable magnetic field may be due to the dynamical relaxation of magnetic twists in the propagating magnetic tower jets .
this causes an increased threshold of the unstable critical wavelength . however , the magnetic tower jets are eventually distorted by the non - axisymmetric perturbations with a few percent of the local sound speed in the system .
the current - driven kink modes grow predominantly on time scales of the order of the local alfvn crossing time @xmath202 . in the gravitationally stratified atmosphere ,
two types of kink modes appears : the internal and external modes . at a large distance away from the central region , the external kink mode grows , while only the internal kink mode exhibits near the central region .
therefore , the exterior of magnetic tower jets will be deformed into a large - scale wiggled structure by the external kink mode
. possible mechanisms of stabilizing / destabilizing in the magnetic tower jets under the gravitationally stratified atmospheres are discussed .
none of the growing surface / body modes of kelvin - helmholtz instability are identified due to the small velocity shear and slow bulk speed ( compared with the local alfvn speed ) .
the pressure - driven instabilities are also inhibited due to the low plasma @xmath182 ( @xmath203 ) inside the magnetic towers .
the `` reversed field pinch '' field configuration in the towers and the external unmagnetized gas may provide some stabilization effects on the current - driven external modes .
non - axisymmetric current - driven instabilities ( @xmath186 ) are absolute instabilities , i.e. , they grow but do not propagate in the jet co - moving ( rest ) frame @xcite .
if the dissipation timescale @xmath204 of the current associated with jets is much longer than the jet dynamical timescale @xmath205 , the patterns created by the cdi modes will persist for some time as the flow advances . and
the bulk flow itself will appear to travel in a true 3-d helical pattern as it follows the magnetic backbone of the helix , as observationally expected in several radio sources as 3c 345 @xcite , 3c 120 @xcite , and 3c 449 @xcite .
helpful discussions with j. finn are gratefully acknowledged .
this work was carried out under the auspices of the national nuclear security administration of the u.s .
department of energy at los alamos national laboratory under contract no .
de - ac52 - 06na25396 .
it was supported by the laboratory directed research and development program at lanl and by igpp at lanl .
appl , s. , & camenzind , m. 1992 , , 256 , 354 appl , s. 1996 , , 314 , 995 appl , s. , lery , t. , & baty , h. 2000 , , 355 , 818 bateman , g. 1978 , mhd instabilities ( cambridge : mit press ) baty , h. , & keppens , r. 2002 , , 580 , 800 baum , s. a. , et al .
1997 , , 483 , 178 begelman , m. c. , blandford , r. d. , & rees , m. j. 1980 , , 287 , 307 begelman , m. c. 1998 , , 493 , 291 benford , g. 1978 , , 183 , 29 benford , g. 2006 , , 369 , 77 bernstein , i. b. , frieman , e. a. , kruskal , m. d. , & kulsrud , r. m. 1958 , proc .
london , ser . a 244 , 17 blandford , r. d. , & payne , d. g. 1982 , , 199 , 883 blandford , r. d. , & znajek , r. l. 1977 , , 179 , 433 bodo , g. , rosner , r. , ferrari , a. , & knobloch , e. 1989 , , 341 , 631 cecil , g. , wilson , a. s. , & tully , r. b. 1992 , , 390 , 365 chiuderi , c. , pietrini , p. , & torricelli - ciamponi , g. 1989 , , 339 , 70 eichler , d. 1993 , , 419 , 111 eilek , j. a. , & owen , f. n. 2002 , , 567 , 202 einaudi , g. , & van hoven , g. 1983 , , 88 , 163 feretti , l. , perley , r. , giovannini , g. , & andernach , h. 1999 , , 341 , 29 frank , a. , lery , t. , gardiner , t. a. , jones , t. w. , & ryu , d. 2000 , , 540 , 342 freidberg j. p. 1982 , rev . mod .
, 54 , 801 gomez , j - l , marscher , a.p . , alberdi , a. , jorstad , s.g . , &
agudo , i. 2001 , , 561 , l161 hardee , p. e. , & rosen , a. 1999 , , 524 , 650 hardee , p. e. , & rosen , a. 2002 , , 576 , 204 hardee , p. e. ( 2004 ) , , 293 , 117 hood , a. w. , & priest , e. r. 1979 , , 64 , 303 hsu , s. c. , & bellan , p. m. 2002 , , 334 , 257 hsu , s. c. , & bellan , p. m. 2003 , , 90 , 215002 hummel , c. a. , et al .
1992 , , 257 , 489 istomin , ya .
n. , & pariev , v. i. 1994 , , 267 , 629 istomin , ya .
n. , & pariev , v. i. 1996 , , 281 , 1 kadomtsev , b. b. , 1966 , rev .
plasma phys . , 2 , 153 kigure , h. , uchida , y. , nakamura , m. , hirose , s. , & cameron , r. 2004 , , 608 , 119 king , i. 1962 , , 67 , 471 kruskal , m. d. , johnson , j. l. , gottlieb , m. b. , & goldman , l. m. 1958 , phys .
fluids , 1 , 421 kudoh , t. , & shibata , k. 1997 , , 474 , 362 lery , t. , baty , h. , & appl , s. 2000 , , 355 , 1201 lery , t. , & frank , a. 2000 , , 533 , 897 li , h. , lapenta , g. , finn , j. m. , li , s. , & colgate , s. a. 2006 , , 643 , 92 ( paper i ) lobanov , a. p. , & zensus , j. a. 2001 , science , 294 , 128 lyubarskii , yu .
e. 1999 , , 308 , 1006 nakamura , m. , uchida , y. , & hirose , s. 2001 , new astronomy , 6 , 61 nakamura , m. , & meier , d. l. 2004 , apj , 617 , 123 nakamura , m. , li , h. , & li , s. 2006 , accepted for publication in apj ; preprint ( astro - ph/0608326 ) ( paper ii ) marshall , h. l. , et al .
2001 , , 549 , l167 meier , d. l. , koide , s. , & uchida , y. 2001 , science , 291 , 84 reid , m. j. , biretta , j. a. , junor , w. , muxlow , t. w. b. , & spencer , r. e. 1989 , , 336 , 112 savolainen , t. , wiik , k. , valtaoja , e. , & tornikoski , m. 2006 , , 446 , 71 shafranov , v. d. 1957 , sov .
phys.jetp , 6 , 545 spruit , h. c. , foglizzo , t. , & stehle , r. 1997 , , 288 , 333 todo , y. , uchida , y. , sato , t. , & rosner , r. 1993 , , 403 , 164 tomimatsu , a. , matsuoka , t. , & takahashi , m. 2001 , , 64 , 123003 vlahakis , n. , tsinganos , k. , sauty , c. , & trussoni , e. 2000 , , 318 , 417 vlahakis , n. , & knigl , a. 2001 , , 563 , l129 vlahakis , n. , & knigl , a. 2003 , , 596 , 1080 zensus , j.a . ,
cohen , m.h . , & unwin , s.c .
1995 , , 443 , 35 | stability properties of `` magnetic tower '' jets propagating in the gravitationally stratified background have been examined by performing three - dimensional magnetohydrodynamic simulations .
the current - carrying , poynting flux - dominated magnetic tower jet , which possesses a highly wound helical magnetic field , is subject to the current - driven instability ( cdi ) .
we find that , under general physical conditions including small perturbations in the initial background profiles , the propagating magnetic tower jets develop the non - axisymmetric , @xmath0 kink mode of the cdi .
the kink mode grows on the local alfvn crossing time scale .
in addition , two types of kink modes appear in the system . at the central region where external thermal pressure confinement is strong
, only the internal kink mode is excited and will grow .
a large distance away from the central region where the external thermal pressure becomes low , the external kink mode is observed . as a result , the exterior of magnetic tower jets will be deformed into a large - scale wiggled structure .
we also discuss extensively the different physical processes that contribute to the overall stability properties of the magnetic tower jets . specifically , when the jet propagates in an initially unperturbed background , we find that they can survive the kink mode beyond the point predicted by the well - known kruskal - shafranov ( k - s ) criterion .
the stabilization in this case comes mainly from the dynamical relaxation of magnetic twists during the propagation of magnetic towers ; the magnetic pitch is reduced and the corresponding k - s critical wavelength becomes longer as the tower jet proceeds .
furthermore , we show that the pressure - driven and kelvin - helmholtz instabilities do not occur in the magnetic tower jets .
this strongly suggests that the cdi is the primary reason for the wiggling structures in jets . |
our knowledge of solar magnetism relies heavily on our ability to detect and interpret the polarization signatures of magnetic fields in solar spectral lines . since the identification of sunspots as regions of strong magnetism by means of the observation of the zeeman effect @xcite ,
the number of physical mechanisms applied to the interpretation of solar spectral line profiles has increased significantly .
the hanle effect observed in linear polarization has been exploited to diagnose weak turbulent magnetic fields @xcite .
a rich spectrum of linear polarization observed near the solar limb named the `` second solar spectrum '' @xcite has been used to constrain magnetic fields and scattering physics .
observations of lines like mn which display hyperfine splitting have been used to break the degeneracy between magnetic flux and field for weak magnetic fields @xcite .
information about plasma kinetics during solar flares is encoded in the linear polarization of h lines through impact polarization @xcite .
other important justifications for multi - line observations exist .
foremost among these is that observations of spectral lines formed at different heights in the solar atmosphere are needed to constrain the magnetic field geometry in three dimensions .
secondly , simultaneous observation of multiple spectral lines and the application of a line - ratio technique have been shown to greatly enhance the diagnostic potential of zeeman - effect observations @xcite . finally , improving technologies ( e.g. ir detector arrays )
have made it possible to take advantage of the increased sensitivity of the zeeman effect with wavelength through the observation of infrared spectral lines @xcite .
it follows that the next generation of stokes polarimeters must have the capability to observe the solar atmosphere in a variety of spectral lines over a wide wavelength range , coupled with the ability to observe several lines simultaneously .
this is reflected in the design of the recently completed spectro - polarimeter for infrared and optical regions ( spinor , @xcite ) , installed at the dunn solar telescope ( dst ) of the national solar observatory on sacramento peak ( nso / sp , sunspot , nm ) , that can observe between 430 and 1600 nm .
post - focus instruments at the recently completed new solar telescope will observe between 400 and 1700 nm @xcite .
the suite of spectro - polarimeters planned for the 4-m advanced technology solar telescope will cover the wavelength range between 380 and 2500 nm @xcite . and
the proposed science for the planned 4-m european solar telescope @xcite emphasizes multi - wavelength observations from the uv to the near - ir .
one immediate instrument requirement stemming from this need for wavelength diversity is that the polarization modulation scheme must be _ efficient _ at all wavelengths of interest .
typically , one attempts to achieve this goal by achromatizing the polarimetric response of a modulator .
this , for instance , is the rational behind the design of super - achromatic waveplates @xcite . in this paper , instead , we present a new paradigm for the design of efficient polarization modulators that are not achromatic in the above sense , since they have polarimetric properties that vary with wavelength .
however , they do so in such a way that they can be operated over a wide wavelength range with near optimal polarimetric efficiency .
we refer to these modulators as _ polychromatic_. because of their performance , these are directly applicable to the next generation of multi - line stokes polarimeters . in the next section , we summarize the basic theory behind polarization modulators .
next , we present example designs using existing technologies , including a discussion of the method used to obtain them .
finally , the performance properties of a prototype polychromatic modulator are presented .
most polarimeters operate by employing retarders and polarizers in a configuration that encodes polarization information into a modulated intensity signal .
this intensity modulation is measured with a detector and analyzed to infer the input polarization state .
an extensive treatment of the theoretical operation of stokes polarimeters has been given in @xcite .
this work draws extensively from that formalism .
the complete polarization state of an input light beam can be described by the stokes vector : @xmath0 where @xmath1 is the intensity , @xmath2 and @xmath3 are the net linear polarizations measured in two coordinate frames that are rotated by @xmath4 with respect to each other , and @xmath5 is the net circular polarization .
since the stokes vector contains four unknown parameters , any polarimeter must make a minimum of four measurements to determine it .
the polarization properties of a polarimeter in any state can be conveniently described by a 4@xmath64 mueller matrix in the usual way .
the first row of any mueller matrix captures the transformation of a stokes vector into intensity . assuming that the detector is sensitive to intensity alone
, then only the first row of the polarimeter mueller matrix is important .
the modulation of intensity by a @xmath7-state polarimeter is captured in the @xmath84 modulation matrix , @xmath9 . for each state , the corresponding row of the modulation matrix ( the _ modulation vector _ )
is given by the first row of the mueller matrix of the polarimeter in that state .
the modulation matrix is generally normalized by the @xmath10 element .
note that , for a retarder - based modulator , the first element of the modulation vector is the same in all modulation states , so after normalization the first column of the modulation matrix is all made of 1s .
the operation of a polarimeter can then be represented by : @xmath11 and the input stokes vector is obtained by @xmath12 where @xmath13 is the demodulation matrix . in @xcite it is shown that , for a given modulation matrix , @xmath9 , the optimal demodulation matrix is @xmath14 .
the polarimetric efficiency of a modulation scheme can be derived from the demodulation matrix itself , and is given by @xcite : @xmath15 where the subscript @xmath16 varies over the four elements of the stokes vector .
the polarimetric efficiency is important in that it quantifies the noise propagation through the demodulation process .
the noise in the @xmath16-th element of the inferred stokes vector , @xmath17 , depends directly on the efficiency as @xcite : @xmath18 where @xmath19 is the uncertainty on the measured intensity , assumed constant , during the measurement cycle .
equation 5 implies that an efficient polarization modulator is required to measure a stokes vector with high precision .
the efficiency on any stokes parameter can be between 0 and 1 , subject to the two independent constraints @xmath20 , and @xmath21 .
many so - called magnetographs are designed to modulate only stokes @xmath5 with high efficiency . in this paper
, we are concerned with modulation schemes that are balanced , in the sense that the efficiencies of stokes @xmath2 , @xmath3 , and @xmath5 are equal . in this case , the maximum ( i.e. , optimal ) modulation efficiency for stokes @xmath1 is 1 , and for stokes @xmath2 , @xmath3 , and @xmath5 , is @xmath22 .
cccc modulator type&retardance&orientation&modulation vector + stepped retarder&@xmath23 & @xmath24 & @xmath25 + & & @xmath26 & @xmath27 + & & @xmath28 & @xmath29 + & & @xmath30 & @xmath31 + lcvrs ( # 1,#2 ) & @xmath32 & @xmath33 & @xmath34 + & @xmath35 & & @xmath36 + & @xmath37 & & @xmath38 + & @xmath39 & & @xmath40 + & @xmath41 & & @xmath42 + & @xmath43 & & @xmath44 + flcs ( # 1,#2 ) & @xmath45 & @xmath46 & @xmath47 + & & @xmath48 & @xmath49 + & & @xmath50 & @xmath51 + & & @xmath52 & @xmath53 + there are several configurations for optimally efficient polarization modulators at a single wavelength .
these may require different numbers of modulation states to achieve the full measurement of the stokes vector .
in table [ tab : monochrome ] , we present examples of optimally efficient and balanced modulators at a single wavelength , employing three different modulator technologies .
these are a rotating waveplate of fixed retardation , a pair of liquid crystal variable retarders ( lcvrs ) with variable retardation and fixed orientation of the fast axis ( e.g. , @xcite ) , and a pair of ferroelectric liquid crystals ( flcs ) with fixed retardance but variable orientation of the fast axis . the particular solution presented here for the flc modulator
is implemented in the diffraction - limited spectro - polarimeter ( dlsp , @xcite ) , which is also installed at the nso / sp dst .
an example of a flc modulator that is optimally efficient but with higher efficiency for circular polarization is given in @xcite .
the lcvr modulator presented here uses 6 modulation states , although it is possible to create a lcvr modulator that is balanced and optimally efficient at one wavelength requiring only 4 states .
we also note that a single lcvr can not modulate all stokes parameters , regardless of the number of modulation states .
modulation efficiency curves for the four stokes parameters between 400 and 1100 nm for modulators with two flcs .
_ continuous curve _ : optimal and balanced solution at 630 nm ( indicated by the central arrow ) corresponding to the nso / dlsp modulator @xcite . _ dotted curve _ : polychromatic solution corresponding to a simple modification of the nso / dlsp modulator , where the first flc is rotated by an additional angle of @xmath54 .
_ dashed curve _ : polychromatic solution obtained from the former solution through the addition of a fixed quartz retarder between the two flcs .
this solution was optimized between 500 and 900 nm ( indicated by the two outmost arrows ) .
the horizontal solid lines in the four panels indicate the maximum theoretical modulation efficiencies that can be achieved simultaneously for the four stokes parameters ( @xmath55 for @xmath1 , and @xmath56 for @xmath2 , @xmath3 , and @xmath5).,width=332 ] the retardance of a given device is wavelength dependent so these standard recipes generally work over a limited wavelength range .
one could change the recipe to access other wavelengths but simultaneous spectro - polarimetric observations in different spectral ranges would not be possible .
the chromatic nature of the example flc modulator is illustrated in fig .
[ fig : flc ] showing the variation of the efficiency with wavelength . in the plot
, we assume a target wavelength of 630 nm and that the intrinsic birefringence of the flcs is not a function of wavelength ( i.e. , no dispersion of the birefringence ) .
we define the region of acceptable performance of a polarimeter as the region over which the efficiency is greater than the optimal efficiency divided by @xmath57 , that is , @xmath58 . at this point , the increase in noise is equivalent to a reduction in the photon flux by a factor of 2 , assuming poisson statistics . by this definition ,
we find that the stepped - retarder solution is efficient over a spectral region of 268 nm , the lcvr solution over 287 nm , and the flc solution over 128 nm .
to create polarimeters that operate over a large wavelength range , instrument developers have typically tried to select optical materials in order to make the polarimeter achromatic ; that is , to make the polarimeter modulation matrix as independent of wavelength as possible .
the success of such an approach depends on the choice of materials and on the desired wavelength coverage . in principle
, this affords an important simplification , which is the possibility of applying a single demodulation scheme to infer the stokes vector at any wavelength . in practice
, the application of a single demodulation scheme can only give a zeroth - order approximation to the true stokes vector , because of the unavoidable residual wavelength dependence of the mueller matrix of the modulator across the spectral range of interest .
for example , the achromatic modulator of nso / spinor shows a 15% variation of its retardance properties over its spectral range of operation ( 430 - 1600 nm , @xcite ) .
therefore one still needs to perform a careful wavelength calibration of the polarimeter in order to infer the true stokes vector within the requirements of polarimetric sensitivity dictated by the science .
previous work @xcite has also attempted to achromatize modulators by adding combinations of fixed and variable retarders , also with the goal of minimizing the wavelength dependence of the resulting mueller matrix . however , the preservation of a given form of the polarimeter s mueller matrix with wavelength is a very limiting , and arguably unnecessary , constraint .
the fundamental driver in the design of a polarization modulator for multi - line applications is the achievement of near - optimal modulation efficiencies in all stokes parameters at all wavelengths of interest .
the mueller matrix of such a modulator can be completely arbitrary , and even strongly dependent on wavelength .
since calibration with wavelength is required , such wavelength dependence will not impact the precision of the polarimetric measurements .
in addition , since the demodulation matrix can be calculated theoretically from the modulator design , it is straightforward to provide real - time approximations of the measured stokes vector at all wavelengths of operation . in designing the polarimeter for the 1-m yunnan solar telescope ( yunnan astronomical observatory , china )
@xcite optimized the efficiency of their polarimeter over a broad wavelength range by adjusting the orientation of the fast axes of the retarding elements , and in our opinion created the first polychromatic modulator .
however , @xcite did not fully optimize their design by allowing the retardations of the elements to vary , they did not adopt an optimal demodulation scheme as described in @xcite which could result in lowering the efficiency that is achieved , and they stressed the importance of minimizing crosstalk which is not necessary to achieve an optimally efficient polarimeter .
we propose a new paradigm where full - stokes polarization modulators are designed to satisfy the constraint of having optimal and balanced polarimetric efficiency at all wavelengths of interest .
this generally results in polarimeters that have modulation and demodulation matrices that are strong functions of wavelength .
we refer to these modulators as polychromatic .
we have developed a technique for optimizing polarization modulators using combinations of fixed and variable retarders .
we find that varying the retardance and orientation of the optical components that comprise a modulator , and maximizing the polarimetric efficiency for all wavelengths of interest , results in realizable configurations with a high degree of wavelength diversity .
we employ an optimization code that systematically searches the parameter space for a solution that maximizes the stokes modulation efficiencies over the desired spectral range .
the search is performed following the strategy of latin hypercube sampling @xcite of the parameter space .
this significantly improves the convergence speed of the search for the optimal solution , compared to a direct monte carlo sampling .
although the solutions found through this method are highly optimized , they may not be unique .
thus it is not possible to conclude with certainty that they represent the solutions with the highest possible efficiency for a given type of modulator .
different optimization schemes based on gradient methods have also been used with success ( e.g. , powell , levenberg - marquardt ) , although they need realistic first guesses in order to converge properly .
we illustrate this concept with three different types of polychromatic modulators for polarimetric applications .
for the first type of modulator , we consider a stack of two flcs , with the possible addition of a fixed retarder to broaden the wavelength range .
for this type of modulator , there are four modulation states .
the broadening of the wavelength range of a flc modulator by the addition of fixed retarders has been suggested @xcite with the intention to minimize off - diagonal elements of the mueller matrix , although that study did not result in efficient modulators over a wide wavelength range . as a first example
, we start with the flc modulator used by the nso / dlsp and presented in table 1 .
the dlsp was designed to operate at 630 nm only .
figure [ fig : flc ] ( continuous curve ) shows that the dlsp , as configured , has optimal and balanced efficiency at 630 nm with a spectral range of acceptable performance of 128 nm .
we then optimized the efficiency of the dlsp modulator between 500 and 900 nm , by allowing the orientation angles of the flcs to vary .
the resulting solution ( dotted curve ) is a simple modification of the nso / dlsp configuration ( cf .
table 1 ) , where the first flc is rotated such that the new switching angles are @xmath54 and @xmath59 . with this new configuration , the modulator still provides optimal and balanced efficiency at 630 nm , however the usable range is now twice as large as before ( 259 nm ) .
next , we added a fixed retarder between the two flcs , and allowed the orientations of the flcs and the retardance and orientation of the fixed retarder to vary .
the resulting solution ( dashed curve ) yields a polarimeter with optimal and balanced efficiency at 630 nm , but the usable range now spans 553 nm .
it must be noted that one can obtain even larger usable ranges with this type of modulator , when the optimization is extended to all the retarding devices .
a clear example of this ( even if still subject to some external constraints ) is the promag modulator illustrated in sect .
[ sec : promag ] . as fig .
[ fig : flc ] , but now for a modulator consisting of a stack of three quartz retarders rotated over 8 discrete steps of 22.5 degrees .
the modulator was optimized between 380 and 1600 nm ( indicated by the two arrows).,width=332 ] the second type of modulator is based on a stack of retarders glued in a fixed set of relative orientations , that is then rotated as a whole to the required set of positions to perform the measurement of the stokes vector .
we consider a stepping modulator , rather than a continuously rotating one .
the adoption of a continuous integration scheme results in a small reduction of the modulation efficiency .
we refer to @xcite for a typical estimation of the reduction of modulation efficiency for continuously rotating modulators . for this type of rotating modulator
, we consider the usual scheme with 8 measurements at positions @xmath60 , with @xmath61 .
the example in fig .
[ fig : ret ] shows the modulation efficiency of a stack of three retarders . as fig .
[ fig : ret ] , but now for a modulator consisting of two lcvrs followed by a fixed quartz retarder , and using a 6-state modulation scheme .
the modulator was optimized between 450 and 1600 nm ( indicated by the two arrows).,width=332 ] the third type is based on a stack of lcvrs followed by a fixed retarder .
the number of states of this kind of modulator is a free parameter .
figure [ fig : lcvr ] considers a modulator with 6 states .
the parameter space for this kind of modulator has a much higher dimensionality than for designs based on flcs or fixed retarders , and therefore multiple solutions with comparable efficiencies are often found . the example shown in fig .
[ fig : lcvr ] was determined by running the optimization several times and selecting the solution with the smallest deviation from optimal efficiency averaged over the optimization range . in the solutions shown in figs .
[ fig : ret][fig : lcvr ] , the fast axis of the first device is conventionally fixed at @xmath62 .
we verified that this constraint does not affect the search of an optimally polychromatic solution for the modulator configuration . for practical purposes
, we also assume that the fixed retarders are made of quartz , while for the lcvrs we assume no wavelength dispersion of the birefringence . while this limits the applicability of the recipes given in this paper , the implementation of realistic birefringence dispersion curves is possible .
preliminary tests show that this is a mildly limiting factor in the search of high - efficiency modulator configurations , typically resulting in a narrower spectral range of optimization for a given number of retarding devices .
specific science applications of polarimeters may require giving a higher priority to selected spectral lines and/or stokes polarization parameters .
for this reason , in our optimization code both the optimizing wavelengths and the four stokes parameters can be attributed different weights in the search of a high - efficiency modulator configuration .
the solutions shown here were derived without any specific application in mind , and therefore they had equal weights for all optimizing wavelengths and stokes parameters . in pratical situations , tweaking of some of the weights may be necessary to correct the behavior of the modulation efficiency in a particular interval of the spectral range .
this is not surprising , especially in the optimization of the modulation efficiency over very large spectral ranges , since one needs to use a correspondingly large number of optimizing wavelengths ( typically , of the order of 50 ) , and the number of nearly equivalent solutions increases dramatically because of the increased number of degrees of freedom in the optimization problem . in such cases ,
the use of different weights can help direct the search towards particular types of nearly equivalent solutions while excluding others , at the user s discretion .
theoretical efficiency curves for the promag modulator , consisting of two flcs followed by a quartz retarder .
the modulator was optimized at 587.6 , 656.3 , 769.9 , and 1083.0 nm ( indicated by arrows ) .
the crosses indicate the measured efficiencies at 587.6 and 1083.0 nm after deployment of the instrument.,width=332 ] as a first application of polychromatic modulators , we re - designed the polarimeter of the hao prominence magnetometer ( promag , @xcite ) .
this instrument was conceived to observe solar prominences and filaments in the spectral lines of he i at 587.6 and 1083.0 nm , and also in h@xmath63 ( 656.3 nm ) .
the re - design was prompted by a series of failed attempts at fabricating the promag modulator following the original design , which was based on a stack of six flcs @xcite .
thus , the original design was replaced by a much simpler scheme , utilizing only two flcs plus a fixed retarder .
the configuration was determined with the optimization method described above and constrained to use one of the flcs from the original design .
retardances of the second flc and the fixed quartz retarder were specified according to the optimized solution , and the retardances of these devices were measured in the laboratory . since the measured retardances differed somewhat from the specified ones , a second optimization was performed varying only the orientations of the acquired devices using their measured retardances .
the modulator was then constructed according to the design resulting from the second optimization .
we estimate that the elements were oriented to the design values to within approximately @xmath64 .
figure [ fig : promag ] shows the predicted efficiencies for the stokes parameters resulting from the optimization .
actual efficiencies at 587.6 and 1083.0 nm were measured after deployment of the instrument at the evans solar facility of the nso / sp , and they are in good agreement with the theory ( see crosses in fig . [ fig : promag ] ) .
the discrepancy observed at 1083.0 nm is likely caused by a less precise determination of the retardances of the modulator optics at that wavelength , possibly due to an undetected leak in the interference filter used during the measurement in the laboratory , combined with the different spectral responses of the promag alpha - nir camera and the photo - diode used in the lab measurement .
theoretical modulation matrix of promag .
the crosses indicate the measured modulation amplitudes at 587.6 and 1083.0 nm .
the first column of the matrix is identically equal to 1 and is not shown.,width=332 ] the theoretical variation of the modulation matrix , @xmath9 , with wavelength is shown in fig .
[ fig : promag_matrix ] .
the first column is unity and therefore is omitted .
measured modulation amplitudes at 587.6 and 1083.0 nm are also shown to be in close agreement with the theory .
in this paper we presented a new paradigm for the design of polarization modulators that responds to the needs of multi - line spectro - polarimetry .
we argued that the effort of achromatizing the response matrix of a polarimeter that is , trying to make the mueller matrix as little dependent on wavelength as possible is both a very limiting and unnecessary constraint .
the only requirement that should be imposed on a polarimeter , in order to reach a given target of polarimetric sensitivity , is that its modulation efficiency be sufficiently high at all wavelengths of interest .
as these wavelengths may be distributed over very large spectral ranges ( of the order of 1000 nm or larger ) , and because of the wavelength dependence of the optical properties of retarding devices , the design of optimally efficient polarization modulators from first principles is a particularly arduous , if not impossible , task . for this reason
, we have developed an improved monte carlo search technique in order to explore the parameter space of a polarimeter , and identify optimally efficient solutions compatible with the particular type of modulator .
we illustrated the power of this method by improving the modulation efficiency of existing polarimetric instruments , as well as by designing new modulators with near optimal and balanced efficiency over very large spectral ranges .
finally we have demonstrated the applicability of this new paradigm by showing the measured performance of a prototype modulator , which was designed and optimized through the method presented in this paper .
we thank d. elmore for helpful discussions during the early stages of this work , s. sewell for helpful comments on the manuscript and frans snik for pointing us to the yunnan polarimeter paper .
the national center for atmospheric research is sponsored by the national science foundation .
a. asensio ramos , m. martnez gonzlez , a. lpez ariste , j. trujillo bueno , and m. collados , `` a near - infrared line of mn i as a diagnostic tool of the average magnetic energy in the solar photosphere , '' * 659 * , 829 - 847 ( 2007 ) .
j. tpn , p. heinzel , and s. sahal - brchot , `` hydrogen h@xmath63 line polarization in solar flares . theoretical investigation of atomic polarization by proton beams considering self - consistent nlte polarized radiative transfer , '' astron . astrophys . * 465 * , 621 - 631 ( 2007 ) .
i. redi , s. k. solanki , w. livingston and j. w. harvey , `` interesting lines in the infrared solar spectrum .
iii . a polarimetric survey between 1.05 and 2.50 @xmath66 m , '' astron . astrophys . * 113 * , 91 - 106 ( 1995 ) .
h. socas - navarro , d. f. elmore , a. pietarila , t. darnell , b. w. lites , s. tomczyk , and s. hegwer , `` spinor : visible and infrared spectro - polarimetry at the national solar observatory , '' solar phys .
, * 235 * , 55 - 73 ( 2006 ) . s. l. keil , t. rimmele , c. u. keller , f. hill , r. r. radick , j. m. oschmann , m. warner , n. e. dalrymple , j. briggs , s. hegwer and d. ren , `` design and development of the advanced technology solar telescope ( atst ) , '' proc .
spie * 4853 * , 240 - 251 ( 2003 ) .
m. collados , `` high resolution spectropolarimetry and magnetography , '' in _ 3rd advances in solar physics euroconference_,b .
schmieder , a. hoffman , j. staude , eds .
. ser . * 184 * ) , pp . 3 - 22 ( 1999 ) .
s. tomczyk , g. l. card , t. darnell , d. f. elmore , r. lull , p. g. nelson , k. v. streander , j. burkepile , r. casini and p. g. judge , `` an instrument to measure coronal emission line polarization , '' solar phys . * 247 * , 411 - 428 ( 2008 ) .
k. sankarasubramanian , b. lites , c. gullixson , d. f. elmore , s. hegwer , k. v. streander , t. rimmele , s. fletcher , s. gregory and m. sigwarth , `` the diffraction limited spectro - polarimeter , '' in _ solar polarization 4 _ , r. casini and b. lites , eds .
. ser . * 358 * ) , pp .
201 - 204 ( 2006 ) .
m. d. mckay , r. j. beckman and w. j. conover , `` a comparison of three methods for selecting values of input variables in the analysis of output from a computer code , '' technometrics * 21 * , 239 - 245 ( 1979 ) .
d. f. elmore , r. casini , g. l. card , m. davis , a. lecinski , r. lull , p. g. nelson and s. tomczyk , `` a new spectro - polarimeter for solar prominence and filament magnetic field measurements , '' proc .
spie * 7014 * , 701416 ( 2008 ) . | information about the three - dimensional structure of solar magnetic fields is encoded in the polarized spectra of solar radiation by a host of physical processes . to extract this
information , solar spectra must be obtained in a variety of magnetically sensitive spectral lines at high spatial , spectral , and temporal resolution with high precision .
the need to observe many different spectral lines drives the development of stokes polarimeters with a high degree of wavelength diversity .
we present a new paradigm for the design of polarization modulators that operate over a wide wavelength range with near optimal polarimetric efficiency and are directly applicable to the next generation of multi - line stokes polarimeters .
these modulators are not achromatic in the usual sense because their polarimetric properties vary with wavelength , but they do so in an optimal way .
thus we refer to these modulators as _ polychromatic_. we present here the theory behind polychromatic modulators , illustrate the concept with design examples , and present the performance properties of a prototype polychromatic modulator . |
geminga ( psrj0633 + 1746 ) is a nearby radio - quiet pulsar , discovered as a gamma - ray source and later identified as a neutron star ( ns ) through optical and x - ray observations ( see @xcite for review ) . the pulsar period ( @xmath6237 ms ) and its derivative ( @xmath7@xmath8 s s@xmath5 ) correspond and @xmath9 from @xmath10 and @xmath11 . ] to a spin - down age @xmath12340 kyr , a spin - down power @xmath13 erg s@xmath5 and a surface magnetic field @xmath14 g. _ hubble space telescope _
parallax measurements confirmed the proximity of geminga ( @xmath15 pc ) and a proper motion corresponding to a transverse velocity of @xmath1210 km s@xmath5 @xcite , exceeding the typical sound speed in the interstellar medium ( 1030 km s@xmath5 ) .
together with crab and vela , geminga is one of known pulsars with the highest spin - down flux ( @xmath16 ) , allowing detailed studies of weak structures in the immediate vicinity of the ns and its interactions with the local interstellar medium .
pulsar wind nebulae ( pwne ) are common diffuse features surrounding nss with very different spin - down ages , that result from the interaction of the relativistic pulsar wind and the ambient medium , producing shocks and outflows that can be observed , mostly because of synchrotron and inverse compton ( ic ) radiation , in a very broad range from radio to gamma - ray energies ( see @xcite ) .
the complex pwn structures mostly depend on the particular wind outflow geometry and on the ratio of the pulsar speed to the sound speed in the ambient medium @xcite .
well - resolved equatorial tori " and axial ( with respect to the spin axis ) jets " are seen around pulsars moving in the interstellar medium with subsonic velocities ( e.g. crab and vela pulsars ; @xcite ) , while in pulsar moving with supersonic speed , as is the case of geminga , bow - like shapes ahead of the pulsar and tails behind are typically observed @xcite .
in fact , _ xmm - newton _ observations of geminga in april 2002 revealed two @xmath12@xmath17-long tails behind the pulsar , approximately symmetric with respect to the sky projection of the pulsar s trajectory @xcite .
geminga was then observed by _
chandra _ in february 2004 unveiling a new structure @xmath125@xmath18-long and @xmath15@xmath18-thick , with a surface brightness @xmath140 times higher than that of the _ xmm - newton _ tails , starting at the pulsar position and aligned with the proper motion direction @xcite .
a faint arc - like structure was also reported 57 arcsec ahead of the pulsar @xcite .
a deeper _ chandra _ observation carried out in august 2007 confirmed the existence of the axial tail ( 50@xmath18-long ) and of the two outer tails discovered by _ xmm - newton _ @xcite providing an overall x - ray pwn luminosity of @xmath19 erg s@xmath5 ( @xmath20@xmath21 ) . comparing 2004 and 2007 _ chandra _ observations , @xcite found indication that the x - ray tail is composed of variable @xmath15@xmath18-long sub - structures ( streaming blobs " ) .
this complex morphology was interpreted as synchrotron radiation possibly arising from the shocked pulsar wind collimated by the ram pressure and/or jets emanating along the pulsar s spin axis .
apart from diffuse tev emission around geminga on larger scales ( @xmath22@xmath23 ) with respect to the x - ray nebula @xcite and a possible detection in the mid - infrared @xcite , the geminga pwn has not been detected in other energy bands so far . in particular , since the discovery of geminga as a gamma - ray source , many observers have attempted to detect it at radio frequencies , both as a continuum and as a pulsating source , providing only tentative detections of weak pulsed emission at low frequencies ( @xmath1100 mhz ; see @xcite ) .
@xcite reported upper limits on the continuum emission ( very large array [ vla ] data ) at 74 mhz ( @xmath2456 mjy ) and 326 mhz ( @xmath245 mjy ) reconciling their nondetection with previous pulsed detections by invoking intrinsic or extrinsic ( refractive interstellar scintillation ) variability . at higher frequencies , @xcite reported an upper limit of 0.5 mjy at 21 cm ( westerbork synthesis radio telescope data ) and 1 mjy at 6 cm ( vla data ) .
more recently , @xcite presented a high - resolution ( 24@xmath18 ) study of the hi interstellar gas distribution around geminga including a 1.4 ghz upper limit of @xmath10.4 mjy in the pulsar direction . here
we present the results of the deepest vla interferometric observation of geminga performed so far .
geminga was observed at 4.8 ghz ( 6 cm ) , with a bandwidth of 50 mhz with the vla in d configuration ( program ap468 ) .
the observation was performed on 2004 july 24 for a total of 6.0 on - source hours , with the antennas pointed at @xmath25 , @xmath26 .
calibration and imaging were performed with the nrao astronomical image processing system ( aips ) software package .
the flux - density scale was calibrated by observing 0137 + 331 ( 3c48 ) .
the source 0625 + 146 was observed at intervals of 30 minutes and used as phase calibrator .
the surface brightness image was produced following the standard procedures : fourier - transform , clean , and restore implemented in the aips task imagr .
we averaged the 2 ifs together in the gridding process under imagr .
self - calibration was applied to remove residual phase variations . in fig .
[ f1 ] we show a @xmath18@xmath17-side field of the radio continuum emission around geminga .
the image has a fwhm beam of @xmath27 ( position angle , @xmath28 ) and a noise level of 25 @xmath29jy beam@xmath5 ( 1@xmath0 ) .
a radio feature , with a peak of brightness at @xmath2210@xmath0 significance level , located at @xmath30 , @xmath31 ( through aips task jmfit ) , has been detected .
the absolute astrometric accuracy of vla observations is linked to the uncertainty in the phase calibrator position , which is better than @xmath32 in our case .
however , the vla accuracy can be limited by a number of effects , such as the atmospheric phase stability .
for this reason , we adopted for the positioning uncertainty a more conservative figure of @xmath33 , which was estimated as the synthesised beam size divided by the signal to noise ratio ( see , e.g. , @xcite ) .
the observed brightness distribution is more extended than that expected for a point source : it is slightly elongated and fits with a single 2-dimensional gaussian model with a fwhm major axis of @xmath34 , a minor axis of @xmath35 , and @xmath36 , incompatible with beam parameters ( the fit with the point - spread function yields a poor reduced @xmath37 of @xmath12.5 for 5 degrees of freedom ) .
the deconvolved size along the major axis ( approximately the direction of geminga proper motion ) is @xmath38 , while it is unresolved in the transverse direction .
the flux density of the source at 4.8 ghz is @xmath39 mjy being calculated by integrating the total intensity surface brightness down to the noise level .
no evidence of source variability was found down to @xmath11.52 hours , which is the minimum time - scale that can be probed in our data . at the distance of geminga ,
the corresponding radio power is @xmath40 w hz@xmath5 . to investigate the presence of such source in other vla observations
, we analyzed an archival dataset at 1.5 ghz in b configuration ( program ak147 ) . in this case
we obtained an image with a fwhm beam of @xmath41 and noise level of 22 @xmath29jy beam@xmath5 ( 1@xmath0 ) .
we did not found any source in correspondence to the 4.8 ghz detection .
@xcite reported on a study of the hi ( 21-cm line ) interstellar gas distribution on a @xmath42 field around geminga based on vla and effelsberg radio telescope data .
based on the line - free channels of the 1.4 ghz cube , they produced a radio continuum image with 24 angular resolution , confirming the nondetection of any source at this frequency , either point - like or extended , down to a noise level of 0.14 mjy beam@xmath5 ( 1@xmath0 ) .
hence , the presently available data do not allow one to constrain the thermal or non - thermal nature of the emission , but , if the radio source is steady , we can set an upper limit on the spectral index , where @xmath43 is the spectral index .
] of @xmath44 ( this value has been calculated by considering that at 1.4 ghz the total flux density is below 3@xmath45 mjy ) .
we note however that if the radio source is variable , this limit is unreliable , since the reported observations are not simultaneous .
we computed the coordinates of the geminga pulsar at the epoch of the reported vla observation by projecting the accurate optical absolute position by @xcite using the proper motion measurement by @xcite : @xmath46 , @xmath47 with errors @xmath2401 .
the pulsar is within the reported diffuse feature , marginally displaced by @xmath3 arcsec from the radio peak emission and in association with a radio brightness of 0.26 mjy beam@xmath5 ( @xmath180% of the peak flux ; fig .
[ f2]).the probability of random association between the radio feature and the geminga pwn region ( approximately fwhm beam size divided by the field of view ) is @xmath24@xmath48 .
we examined radio , optical and near - infrared catalogs looking for possible background objects of different nature ( e.g. an agn jet ) , but no counterpart compatible with the observed structure was found . in order to compare the radio and x - ray morphologies of the structures associated to geminga
, we retrieved a public _
acis - s imaging exposure of the field performed close to the vla observation ( obs .
i d 4674 ; @xcite ) .
the @xmath120-ks - long observation was carried out on 2004 february 07 ( 168 days before the vla observation ) in ` timed exposure ' mode using a 1/8 subarray .
geminga was positioned in the back - illuminated acis - s3 chip , sensitive to photons between 0.2 and 10 kev .
the data were processed using the _ chandra _ interactive analysis of observation software ( ciao , version 4.2 ) and we employed the caldb 4.3 calibration files .
standard screening criteria were applied to generate a 0.58 kev image of the field of geminga . using the _ chandra _ aspect tool , we found a small offset in the astrometry ( @xmath49 , @xmath50 ) . after fixing the offset , in order to assess the absolute astrometry of the _ chandra _ data set , we computed the position of geminga with the wavedetect task .
we localized it at @xmath1@xmath51 from the position expected on the basis of the source absolute optical position and proper motion , well within the typical acis - s localization accuracy .
the x - ray contours obtained from the 2004 _ chandra _ observation and related to the pulsar and to the axial pwn tail ( see also @xcite ) are shown in fig .
the deconvolved length of the radio feature ( @xmath110 ) is smaller than the whole time - integrated ( 2004 and 2007 _ chandra _ combined data ) x - ray tail , but larger than the substructures ( typically @xmath15-long ` blobs ' ) detected by _
chandra _ in individual observations @xcite .
a vla observation at 4.8 ghz unveiled a @xmath1@xmath2-long radio feature with flux of @xmath10.4 mjy associated with the geminga pulsar and its pwn . significant contamination from geminga persistent pulsed radio emission
is very unlikely : given the tight upper limits on pulsed flux density at low frequency ( e.g. @xmath52.05 mjy at 0.4 ghz ; @xcite ) , the contribution of pulsed emission to the total observed flux at 4.8 ghz would be @xmath240.1% for the mean spectral index of radio pulsars ( @xmath24@xmath43@xmath53 ; @xcite ) and could at most reach @xmath115% even in the extreme case of a flat spectrum . the occurrence of transient pulsed ( or continuum ) radio emission in isolated nss in not unheard - of , but it involves at our present knowledge pulsar classes showing a totally different phenomenology from that of the geminga pulsar , like magnetars and rotating radio transients @xcite .
furthermore , the emission models mentioned above , as well as any other kinds of magnetospheric emission originating very close to the ns surface ( and hence appearing as point - like at the spatial resolution of the available observations ) , are disfavored by the evidence for a slightly elongated feature , together with the occurrence of a marginal offset between the ns position and the radio emission centroid ( @xmath54 ) .
the latter two features and the tendency of the radio emission to distribute along the direction of the axial x - ray tail , naturally lead to an association with the geminga pwn .
* and references therein ) observed that the axial tail may result from the shocked pulsar wind collimated by the ram pressure ( i.e. a bow - shock ) or from a jet emanating along the pulsar s spin axis and perhaps aligned with the direction of motion . in the frame of the standard bow - shock models ,
radio pwne are generally interpreted as arising from cooling of x - ray - emitting leptons @xcite .
the smaller size of the radio nebula of geminga ( with respect to the whole , time - integrated , @xmath150 long x - ray tail ) apparently contradicts this hypothesis , unless we associate the observed radio feature to the variable streaming ` blobs ' of @xmath15 diameter , possibly constituting the x - ray axial tail as observed in 2004 @xcite .
furthermore , synchrotron models explaining the x - ray tails as pulsar wind shocked by the ram pressure due to the ns s supersonic motion typically require relatively low magnetic fields ( @xmath110100 @xmath29 g ; @xcite ; @xcite ) . with such low magnetic fields
, the radio emission arising from the cooling of the x - ray emitting electrons , would be placed well outside the vla field of view due to the large proper motion of geminga : e.g. more than 0.5 deg off the position of the x - ray emission for a magnetic field of 100 @xmath29 g . in summary ,
ram - pressure collimated pwn models can be applied to our case only if additional processes , such as adiabatic expansion from hypothetical over - pressurized substructures of the tail ( not observable at the present stage ) provide a cooling power much higher than the synchrotron one and/or if radio emitting electrons are directly accelerated at the pwn shock , i.e. they are not originated from the cooling of high - energy electrons as in standard pwn models . in this respect ,
we note that the radio wisps " observed in the crab pwn are produced by the same process as the higher - energy ones , although the mechanism by which this particle spectrum is generated is still not known ( see e.g. @xcite ) .
alternatively , the radio tail of geminga could in principle be ascribed to a ( magnetically - collimated ) jet , in which the radio emission could directly result from radio - emitting electrons freshly - injected along the spin - axis at the jet formation site , or from internal shocks in which the accelerated x - ray - emitting leptons cool down to radio energies on the observed short space and time scales ( see e.g. @xcite for possible models of magneto - hydrodynamic jet produced by isolated pulsars ) . in the latter case ,
magnetic fields @xmath550.05 g are required in order to avoid the occurrence of an additional ( i.e other than synchrotron ) cooling process . as already noticed for the case of the ram - pressure collimated pwn models ,
there is no evidence of such high fields in other pwne . however , in the first place , large magnetic fields due to the ns dipole are in principle available at the pulsar light cylinder ( @xmath56 g for geminga ) and in its surroundings . in the second place
, an observational bias may also be at work here .
in fact , despite the growing number of jet - like structures around nss resolved in x - rays @xcite , no ` genuine ' magneto - hydrodynamic radio jet around pulsar was detected so far ( see e.g. @xcite for vla mapping of numerous pulsar fields ) , possibly excepting psrj2021 + 4026 , another radio - quiet gamma - ray pulsar , located at the edge of the supernova remnant g78.2 + 2.1 @xcite . thus , it can not be excluded that pulsar radio jets might be observable only in nearby and radio - quiet pulsars , not hindering their close environment by bright magnetospheric radio emission and having magnetic fields much stronger than those related to other pwn structures placed at larger distances from the ns surface .
interestingly enough , this hypothesis may be checked by a series of deep radio interferometric exposures . scaling our result for geminga
, this survey could unveil jet - like structures in few other near ( @xmath571 kpc ) and energetic ( @xmath58 erg s@xmath5 ) radio - quiet pulsars , some of them recently discovered by the fermi satellite @xcite .
finally , we note that a further possible explanation of the geminga radio tail is the leakage of radio - emitting electrons ( with lorentz factor @xmath59100 ) directly from the open field lines of the magnetosphere .
the maximum current through the magnetospheric accelerator ( i.e. , the goldreich - julian current ; see e.g. @xcite and references therein ) and then the maximum particle flux escaping from geminga open field lines can be estimated from @xmath60 s@xmath5 ( where @xmath61 is the pulsar angular velocity , and the ns s radius @xmath62 is assumed to be 10 km ) .
the expected synchrotron power at 4.8 ghz , for a random distribution of pitch angles , can be roughly obtained from : @xmath63 where @xmath64=@xmath65@xmath66/@xmath67 is the synchrotron cooling time of the electrons , @xmath68 is the magnetic field energy density and @xmath69 is the fraction ( w.r.t .
@xmath70 ) of radio emitting particles .
the observed luminosity of 10@xmath71 erg s@xmath5 at 4.8 ghz would then be matched for @xmath72 , implying a magnetic field of @xmath73 g in the area of the emission .
if this magnetic field is associated to the ns dipole , the emitting region would then be placed at a distance of @xmath74 cm from the ns surface . in summary , for @xmath75 ( in turn implying @xmath76 of the order of ten light cylinder radii , @xmath77 , @xmath78 g , and @xmath79 yrs ) , this model would provide a cooling and possibly expanding diffuse radio feature protruding from the geminga pulsar .
if this model is correct , two geometrical constraints should be simultaneously satisfied by the geminga pulsar : the radio pulsar beam not intersecting the line of sight and the leaking electrons producing a relatively narrow cone of radio emission roughly aligned with the projection of the pulsar proper motion .
this would suggest that geminga is a nearly aligned rotator , with the spin axis almost aligned with the proper motion and with the velocity vector close to the plane of the sky .
we have reported on the discovery of a radio emitting structure with a flux of @xmath10.4 mjy at 4.8 ghz , associated with the geminga pulsar and roughly aligned with the direction of the neutron star proper motion .
the source appears slightly extended ( @xmath1@xmath2-long ) but further deeper observations are needed to confirm the elongation unambiguously .
the nature of this emission is unconstrained by present data , but a combination of dedicated radio interferometric and x - ray observations is expected to shed light on the underlying physical process and likely unveil similar radio tails in other radio - quiet pulsars .
vla is operated by the national radio astronomy observatory ( nrao ) , a facility of the national science foundation ( nsf ) operated under cooperative agreement by associated universities , inc .
this research has also made use of data obtained from the _ chandra _ data archive and software provided by the _
chandra _ x - ray center .
p.e . acknowledges financial support from the autonomous region of sardinia through a research grant under the program po sardegna fse 20072013 , l.r .
7/2007 , `` promoting scientific research and innovation technology in sardinia '' .
the authors are grateful to the anonymous referees , whose valuable comments led to substantial improvements in the paper .
v. m. , roberts m. s. e. , harding a. k. , 2006 , isolated neutron stars , in compact stellar x - ray sources .
cambridge astrophysics series , no .
lewin , w. h. g. & van der klis , m .. cambridge , uk : cambridge university press a. , mattana f. , de luca a. , mereghetti s. , caraveo p. , conti m. , tavani m. , 2005 , in f. a. aharonian , h. j. vlk , & d. horns ed . , high energy gamma - ray astronomy vol .
745 of american institute of physics conference series , melville ny , pp 371376 | a deep very large array observation of the geminga pulsar field led to the discovery , at a higher than 10@xmath0 significance level , of radio emission trailing the neutron star proper motion .
this @xmath1@xmath2-long radio feature , detected with a flux of @xmath10.4 mjy at 4.8 ghz , is marginally displaced ( @xmath3 arcsec ) from the pulsar ( which , at any rate , is unlikely to contribute with magnetospheric pulsed emission more than 15% to the total observed radio luminosity , @xmath1@xmath4 erg s@xmath5 ) and positionally coincident with the x - ray axial tail recently discovered by _
chandra _ and ascribed to the pulsar wind nebula ( pwn ) .
overall , the geminga radio tail is compatible with the scenario of a synchrotron - emitting pwn , but the present data do not allow us to discriminate between different ( and not always necessarily mutually exclusive ) possible processes for producing that .
if this radio feature does not result from intrinsic peculiarities of geminga , but its proximity and radio - quiet nature ( both helping in not hindering the faint diffuse radio emission ) , other relatively near and energetic radio - quiet pulsars could show similar structures in dedicated interferometric observations .
[ firstpage ] pulsars : general pulsars : individual ( j0633 + 1746 , geminga ) radio continuum : stars stars : neutron stars : winds , outflows . |
the matter bounce scenario is an alternative to inflation where scale - invariant perturbations are generated in a contracting cosmology , which is connected to our expanding universe via a non - singular bounce . to be specific , during a phase of matter - dominated contraction ( i.e.
, when the matter field dominating the dynamics has vanishing pressure ) , the fourier modes of the co - moving curvature perturbation that exit the sound horizon become scale - invariant , assuming they were initially vacuum quantum fluctuations . for a more detailed introduction to the matter bounce scenario ,
see the review @xcite .
the simplest realizations of the matter bounce scenario are those where the matter field is assumed to be a scalar field with an appropriate potential so that the pressure vanishes ( at least on average ) @xcite .
however , there are two important differences between observations and the predictions of this family of matter bounce scenario .
first , this type of realization of the matter bounce scenario predicts an exactly scale - invariant spectrum , rather than the observed slight red tilt with a spectral index of @xmath0 @xcite , and second , the tensor - to - scalar ratio @xmath1 is predicted to be significantly larger than what is allowed by the observational bound @xmath2 @xcite . for these reasons ,
this type of matter bounce scenario with a single scalar field is ruled out by observations @xcite .
just as there are many inflationary models based on scalar fields with different potentials or various modified gravity theories , there are also a number of realizations of the matter bounce scenario , some of which predict a slight red tilt in the spectrum of scalar perturbations and also a sufficiently small tensor - to - scalar ratio .
a slight red tilt can be produced if the equation of state of the matter field is slightly negative @xcite , and there are three known mechanisms for predicting a smaller tensor - to - scalar ratio : ( i ) including additional matter fields , in which case entropy perturbations become important and can increase the amplitude of the scalar perturbations without affecting the tensor modes @xcite , ( ii ) choose a matter field that has a small sound speed , which increases the amplitude of vacuum fluctuations of the scalar perturbations @xcite , and ( iii ) suppress the tensor - to - scalar ratio during the bounce due to , e.g. , quantum gravity effects as has been found to occur in loop quantum cosmology @xcite .
moreover , the energy scale of the universe during the contracting pre - bounce phase can be comparable to the one of the present universe @xcite . based on the underlying idea of effective field theory that relevant degrees of freedom are determined by the energy scale , it is natural to postulate that the low energy degrees of freedom today are also likely to be the appropriate degrees of freedom during the matter contraction stage in the pre - bounce phase . in this case
, the matter content in the contracting universe would be the same as what is observed today , i.e. , dark matter and dark energy .
interestingly , these considerations do in fact influence the generation of primordial perturbations in a way which naturally avoids the problems of the simplest realizations of the matter bounce scenario described above .
first , due to the presence of dark matter and dark energy , there will exist an era when the dark matter dominates the dynamics of the background , but that dark energy also contributes slightly . at this time
, the effective equation of state will be slightly negative , and this will produce almost scale - invariant perturbations with a slight red tilt .
second , cold dark matter is known to have a small sound speed @xcite and this will increase the amplitude of the vacuum quantum fluctuations of the scalar perturbations , which in turn will predict a smaller tensor - to - scalar ratio .
thus , a matter bounce scenario based on the matter content observed in the universe today can naturally predict scale - invariant perturbations with a slight red tilt and a small tensor - to - scalar ratio @xcite .
further properties of such a matter bounce scenario with dark matter and dark energy , as well as some extensions thereof , have been studied in @xcite .
bouncing cosmologies with dark matter and dark energy clearly necessitate physics beyond the standard model . for a bounce to occur
, the singularity theorems of general relativity must be avoided , and this requires either modifications to the einstein equations ( e.g. , due to quantum gravity effects , or in @xmath3 theories ) or the presence of a matter field which violates the null energy condition , depending on the specific realization of the matter bounce scenario .
one interesting possibility reviewed below in sec .
[ ss.fermi ] is that fermionic matter can potentially reproduce both dark matter and dark energy as well as generate the bounce ( although it has not yet been shown that it can do all three simulatenously ; this is a question for future work ) .
another possibility is to describe dark matter and dark energy in a phenomenological manner as perfect fluids with some appropriate properties , this is often done if the bounce is assumed to occur due to modifications to the einstein equations .
it is important to note that the main qualitative predictions of the dark matter bounce scenario reviewed appear to be independent of the specific realization , including a positive running of the scalar spectral index which allows this model to be differentiated from inflationary and ekpyrotic cosmologies @xcite .
the quantitative predictions do in fact depend on these details , and these can be used to rule out some specific realizations of the dark matter bounce scenario @xcite . in this paper , we will describe the matter bounce scenario with dark matter and dark energy in some detail , and in particular review its main predictions . in sec .
[ s.lcdm ] , we will consider the simplest such cosmology where dark energy is due to a cosmological constant @xmath4 and cold dark matter can be treated as a perfect fluid with a constant sound speed . while this scenario does indeed predict a nearly scale - invariant power spectrum and a small tensor - to - scalar ratio , it also predicts a large positive running of the scalar index which rules it out .
this shows that for the matter bounce scenario based on dark matter and dark energy to be viable , some form of interacting dark matter and/or dark energy is required ; one particularly interesting interacting dark energy model is reviewed in sec .
[ s.int ] .
then , in sec . [ s.obs ]
we derive some generic predictions of this family of matter bounce scenarios that can be used to differentiate it from inflationary , ekpyrotic , or other cosmological scenarios , including the running of the scalar index and non - gaussianities . finally , in sec .
[ s.bounce ] , we describe different physical processes that can produce a bounce , either by modifying the einstein equations or with matter fields violating the null energy condition .
we end with a discussion in sec .
[ s.disc ] .
in the @xmath4cdm bounce scenario , dark energy is assumed to be due to a small cosmological constant @xmath4 and dark matter is treated as a perfect fluid with a constant equation of state ( assuming a constant equation of state ) and solutions to the mukhanov - sasaki equation have a singular limit when @xmath6 . ]
@xmath7 , which is taken to be positive and very small compared to 1 .
furthermore , it is assumed that radiation is also present and this will dominate the cosmological dynamics when the radiation fluid becomes sufficiently dense . for a more detailed presentation of this cosmological scenario ,
see @xcite . assuming the background space - time to be the spatially flat friedmann - lematre - robertson - walker ( flrw ) cosmology with the line element ds^2 = a()^2 , in the contracting branch of this cosmological model the classical friedmann equation is given by [ friedmann ] h^2 = ( _ d + _ m + _ ) , where the dark energy density @xmath8 is constant , the dark matter energy density @xmath9 and the radiation energy density @xmath10 . in the contracting branch ,
the cosmological dynamics will first be dominated by the cosmological constant , then by the cold dark matter , and finally by radiation . in this way ,
given the pressures @xmath11 and @xmath12 , the effective equation of state is given by [ weff ] ( ) , and runs monotonically from @xmath13 to @xmath14 , as shown by the dashed black line in fig .
[ rafig ] ( although the figure ends at a time of matter - domination when @xmath15 ) .
in particular , during the matter contraction stage , for some time the effective equation of state will be negative and very close to 0 , i.e. , @xmath16 , with @xmath17 .
the fourier modes of the perturbations that exit their sound horizon at this time will be nearly scale - invariant , with a slight red tilt .
to see this , recall that in cosmological perturbation theory , the final scale - dependence of the power spectrum essentially depends on the effective equation of state at their time of horizon - crossing .
this is because the fourier modes inside the horizon oscillate adiabatically and do not feel the space - time curvature , while modes outside the horizon either freeze ( in an expanding space - time ) or all grow at the same rate ( in a contracting space - time ) .
so , the scale - dependence is determined by the dynamics of the background space - time at horizon - crossing , which in turn depend on the effective equation of state of the matter fields .
now , consider the fourier modes that exit their sound horizon when the effective equation of state is given by @xmath16 .
for these modes , it is reasonable to assume that the effective equation of state is constant ( which is a good approximation during the time of horizon crossing , which is when the background dynamics is most relevant ) , and so for these modes the scale factor can be approximated to be [ scalef ] a ( ) = a_o ^2(1 + 3 ) , where terms of order @xmath18 and smaller have been dropped .
note that @xmath19 for a contracting flrw space - time .
recall that the dynamics of scalar perturbations on a spatially flat flrw background space - time are determined by the mukhanov - sasaki equation , which in fourier space is [ eom ] v_k
+ c_s^2 k^2 v_k - v_k = 0 , where primes denote derivatives with respect to @xmath20 , @xmath21 , @xmath22 is the sound speed of cold dark matter , and the mukhanov - sasaki variable @xmath23 is related to the comoving curvature perturbation @xmath24 by @xmath25 . for a detailed review of cosmological perturbation theory , see , e.g. , @xcite .
therefore , in the regime where holds the effective equation of state can be approximated to be constant does evolve ( although slowly ) and @xmath26 , there are additional terms coming from the @xmath27 term in the mukhanov - sasaki equation of the form @xmath28 and @xmath29 that are of the same order as @xmath30 .
we have dropped these terms here since this model is simply used for illustrative purposes and is in any case observationally ruled out . in later sections where @xmath31 evolves significantly more slowly due to interactions between dark matter and dark energy , then
the terms coming from @xmath32 and @xmath29 in @xmath27 are subleading and can safely be neglected when @xmath33 . ] , and since the sound speed @xmath34 is constant the mukhanov - sasaki equation in this regime becomes ( again , dropping terms of order @xmath18 and smaller ) v_k
+ c_s^2 k^2 v_k - v_k = 0 . assuming vacuum quantum fluctuations as the initial conditions at @xmath35 , the solution is given by v_k = h_n^(1)(-c_s k ) , where @xmath36 is the hankel function with @xmath37 .
once the fourier mode has exited the sound horizon , i.e. , for @xmath38 , the asymptotics of the hankel function show that @xmath39 grows as v_k ~ , and therefore the scalar perturbations that exit the sound horizon when @xmath16 have a nearly scale - invariant spectrum with a slight red tilt @xcite , _
k ~ , with the scalar spectral index n_s - 1 = - 12 . clearly , the amplitude of the perturbations will continue to increase until the radiation fluid begins to dominate the background dynamics , at which point @xmath40 and the perturbations will no longer be amplified .
therefore , it is the hubble rate at the time of equality between @xmath41 and @xmath42 that determines the amplitude of the scalar perturbations .
note that the sound speed also affects the amplitude of the scalar perturbations , and that the smaller the sound speed is , the larger the amplitude of the scalar perturbations becomes .
this effect is only present for scalar perturbations since the sound speed of tensor perturbations is always 1 .
therefore , for a small sound speed , the scalar perturbations are amplified compared to the tensor perturbations . in this way
, a small sound speed leads to a ( potentially much ) smaller tensor - to - scalar ratio than what would be expected otherwise .
another important point here is that the above analysis holds for the modes that exit the sound horizon when @xmath16 .
however , since different modes exit at different times and the effective equation of state evolves with time , the scalar spectral index will run .
a simple calculation shows that the running spectral index of curvature perturbation _
s = > 0 , which follows from @xmath43 and @xmath44 , the second relation being a consequence of the effective equation of state increasing as modes with larger wave numbers @xmath45 ( or , equivalently , shorter wavelengths ) exit the sound horizon @xcite .
this positive running of the spectral scalar index is a generic prediction of bouncing cosmologies based on dark matter and dark energy , and offers a clear way to discriminate between these models and other cosmological scenarios like inflation and the ekpyrotic universe which both typically predict a negative running of the scalar spectral index @xcite .
in fact , in the case where dark energy is due to a cosmological constant and dark matter is treated as a perfect fluid with a constant equation of state , it is possible to calculate the running of the scalar spectral index explicitly @xcite . during the matter - dominated era of contraction , neglecting the radiation field and making the approximation that [ weff2 ] = - , it follows that @xmath46 since @xmath47 .
furthermore , the fourier mode @xmath45 exiting the sound horizon at any given time is determined by the relation @xmath48 and therefore [ keta ] = - .
combining these results gives @xmath49 using the observed value of @xmath50 @xcite .
this amplitude of the running of the scalar spectral index is larger than what is allowed by the latest observations , which constrain @xmath51 to be at most of the order of @xmath52 @xcite , and therefore this simplest @xmath4cdm bounce scenario is ruled out observationally
. however , there are other realizations of a bounce scenario based on dark matter and dark energy , some of which predict a smaller running of the spectral scalar index .
indeed , interactions between dark matter and dark energy can potentially decrease the rate at which the effective equation of state grows and hence give a smaller ( though still positive ) @xmath51 . as a final comment , note that observational constraints rule out any bouncing cosmological scenario based on dark matter and dark energy with a symmetric bounce .
the reason for this is that in the contracting branch the fourier modes must exit at the time when dark matter dominates for the resulting spectrum to be nearly scale - invariant ; however , most of the modes observed today in the cosmic microwave background ( cmb ) reentered the horizon in the expanding branch at a time of radiation - domination .
therefore , the bounce must be significantly asymmetric , with a longer radiation - dominated era in the post - bounce expanding branch of the cosmology , for the dark matter bounce scenario to be viable .
( in fact , this is a constraint that all bouncing scenarios must satisfy , not just this matter bounce scenario . ) for the case of the dark matter bounce scenario , this asymmetry could be generated by interactions between dark matter and dark energy .
dark energy and dark matter are the two dominant components that govern the evolution of the universe at present ( see @xcite for comprehensive reviews ) . in the standard cosmological paradigm ,
i.e. , the @xmath4cdm model , these two sectors do not interact with any other matter fields ( or each other ) except via gravitational effects .
however , small interactions between dark energy and dark matter are not ruled out and could potentially play an important in the evolution of the universe ; models that include interactions between dark energy and dark matter are typically called interacting dark energy models @xcite . from the theoretical prespective , models of this type can alleviate the cosmic coincidence problem of why the dark energy and dark matter energy densities today have the same order of magnitude .
furthermore , the recent boss experiment @xcite indicates a slight deviation ( at @xmath53 cl ) in the expected @xmath4cdm value of the hubble parameter and the angular distance at an average redshift of @xmath54 , which can successfully be modeled by interacting dark energy @xcite . in the interacting dark energy models , the background evolution of the universe is still described by the friedmann equation , but now the equations of motion for the matter fields are modified .
as is well known , the energy - momentum tensor for a matter field that does not directly interact with any other matter fields satisfies @xmath55 .
however , if interactions are present the equations above must be appropriately modified to _ t^_i = q^_i , where @xmath56 describes the energy - momentum transfer between different components . here
we are interested in the case where there are interactions between dark energy and dark matter . in this case , the combined energy - momentum tensor for the dark sector is conserved ( i.e. , we assume there are no interactions with the radiation field ) , giving [ int - t ] _ t^_d+_t^_m= q^_d+q^_m=0
. therefore , the energy transfer satisfies @xmath57 : energy flows from dark energy to dark matter for @xmath58 , and the flow is reversed for @xmath59 .
specific realizations of the dark sector energy - momentum tensor have been studied in some detail for the cases where the dark sector is composed of a yang - mills condensate @xcite , or of an invisible sector of quantum chromodynamics @xcite . for our purposes in this paper
, we will consider the simple model where @xmath60 , with @xmath61 being a constant , and the equation of state of dark energy @xmath62 is a constant allowed to slightly differ from -1 .
( in fact , @xmath62 should be slightly smaller than @xmath13 in order to ensure stability of the curvature perturbations @xcite . ) then , the energy conservation equations for the dark sector become [ dm ] + 3(1-)=0 , [ de ] + 3(1+w_d)+3=0 .
the scale factor @xmath63 , which is a monotonic function of time in the contracting pre - bounce phase , is used as a relational ` time ' here .
it is convenient to define the ratio of matter and dark energy densities as @xmath64 .
then , in the limit where the radiation field is negligible and the pressure of dark matter is assumed to vanish , the effective equation of state can be expressed as = .
the equation of motion for @xmath65 follows from and , [ ra ] = , and can be solved analytically .
choosing initial conditions such that at the relational time @xmath66 the ratio of densities has the value @xmath67 , the solution is [ ras ] ( a)= .
the two asymptotic solutions , for @xmath68 and @xmath69 , are _ a0= - _ a+= 0 .
therefore , at early times in the contracting branch @xmath65 is vanishing and @xmath70 .
then , as the space - time contracts , the scale factor decreases and @xmath65 increases and asymptotes to @xmath71 , at which point @xmath72 .
this process is shown by the solid blue line in fig .
[ rafig ] .
( at some later time the radiation field will become important , but here we are most interested in the regime where the dark sector is driving the dynamics of the space - time . ) in @xmath4cdm and interacting dark energy models .
note that time goes from left to right in this plot for a contracting universe . here
we set @xmath73 and @xmath74 for demonstration . ]
now it is possible to calculate the scalar power spectrum generated by this background space - time for the fourier modes that exit their sound horizon when the effective equation of state @xmath31 is nearly vanishing , but slightly negative .
in fact , the calculations in sec .
[ ss.lcdm-p ] are easily adapted to these background dynamics ; the only difference is that the effective equation of state ( denoted by @xmath75 in sec . [ s.lcdm ] ) evolves more slowly in this case where there are interactions between dark matter and dark energy . assuming that the fourier modes of interest exit the sound horizon at a time where @xmath65 is near to its asymptote , the scalar spectral index for these perturbations is given by , as before , @xmath76 , and since @xmath72 n_s = 1 - 12 .
the most recent measurements by the planck collaboration give @xmath77 @xcite , which constrains the strength of the coupling constant @xmath78 in the interacting dark energy model .
interestingly , this value is compatible with constraints provided by late time observations .
for instance , an analysis of late - time effects in the temperature - temperature planck data combined with polarization data from the 9-year results of the wilkinson microwave anisotropy probe and measurements of baryon acoustic oscillation measurements found @xmath79 at the 68% confidence level , and @xmath80 ( again at the 68% confidence level ) if type ia supernovae data and the latest measurement of the hubble rate by the hubble space telescope are also included @xcite .
given the uncertainties in these constraints , these values are in reasonable agreement with the @xmath81 required to provide the observed tilt in the scalar power spectrum for the dark matter bounce scenario .
it is also straightforward to extend the calculation of the running of the spectral scalar index to the interacting dark energy model . as for the calculation in for the @xmath4cdm bounce , the relations @xmath43 and @xmath82 still hold true . on the other hand , the evolution of @xmath31 is different : using eq
gives for the case that @xmath83 , and in the matter contraction stage where @xmath84 and @xmath85 where @xmath86 , this approaches [ dweff ] = -^2 , giving a running of the spectral index of [ alpha - int ] _ s= ( n_s-1)^2 . for @xmath50 ,
the interacting dark energy model studied here predicts @xmath87 , which is consistent with the latest observational constraints , namely that @xmath88 be at most of the order of @xmath52 @xcite .
importantly , this positive running provides a distinguishable feature of the matter bounce scenario , since typical inflationary and ekpyrotic cosmologies predict @xmath88 to be negative @xcite .
this is slightly favored by the latest data analysis which suggests a small positive running of the scalar spectral index @xcite .
it is possible to calculate higher orders of the running in this model , for example the running of the running .
the running of the running index is defined as _
s , which can easily be calculated by rewriting @xmath89 , with the second term being precisely @xmath88 ; the result is _
s = ( n_s-1)^3 , which for @xmath50 gives @xmath90 .
in addition to being consistent with the latest observational results , a bouncing cosmology with dark matter and dark energy also gives novel predictions that can be used to distinguish this scenario from other early universe models like inflation @xcite .
one of the key predictions of a bouncing cosmology with dark matter and dark energy is that the spectrum of the primordial curvature perturbations generated will have a positive running @xcite .
this can be seen explicitly above in and for the specific realizations based on @xmath4cdm and interacting dark energy presented in the previous sections .
importantly , this prediction is independent of the specific properties of the dark matter and dark energy ; rather , it is a consequence of the fact that since , roughly speaking , @xmath91 and @xmath92 , then as the universe contracts the dark matter contribution to the background energy density becomes more and more dominant .
therefore , the effective equation of state increases monotonically during this process , as can be seen from . furthermore , since the fourier modes of the cosmological perturbations with larger wavenumbers @xmath45 ( i.e. , shorter wavelengths ) exit their sound horizon at later times than modes with short wavelengths , it follows that @xmath93 . from this , it follows immediately that that the running of the scalar spectral index @xmath94 is necessarily positive . on the other hand , in typical slow - roll inflation scenarios ,
the running is negative .
this is because during inflation , the hubble rate slowly decreases as the inflaton rolls down the potential , and it is this decrease in the hubble rate that generates a red tilt in the spectrum of curvature perturbations .
furthermore , the hubble rate decreases at an increasingly rapid rate ( as is required for a graceful exit from inflation ) , giving a redder spectrum for the modes that exit their horizon at later times , i.e. , for larger @xmath45 ; this corresponds to a negative running of the scalar spectral index .
( it is possible to generate a positive running in inflationary scenarios , however this requires large modulations in the potential of the inflaton field @xcite . ) moreover , ekpyrotic cosmologies also predict a negative running of the scalar spectral index @xcite .
for this reason , a precise measurement of @xmath88 could differentiate between these scenarios .
moreover , as shown above , specific realizations will provide specific quantitative predictions for @xmath88 as well as higher order parameters .
for example , the interacting dark energy model studied in sec .
[ s.int ] makes a clear prediction for all of the higher order terms @xmath95 ( where @xmath96 and @xmath97 ) , clearly , running parameters in this case decrease by one power of @xmath98 at each order .
this consistency relation between the scalar spectral index and the running parameters is a clear signature for the matter bounce with the interacting dark energy model of sec .
[ s.int ] .
the equation of motion for the tensor perturbations in fourier space is _ k
+ k^2 _ k - _ k = 0 , which is quite similar to the equation for scalar modes , especially for the case considered here where @xmath31 is evolving so slowly that @xmath32 and @xmath29 are negligible compared to @xmath31 : in this case @xmath99 ( in the limit that @xmath32 and @xmath29 terms are neglected ) and the equations of motion are nearly identical for the scalar and tensor perturbations .
however , there is one important difference between the two equations of motion : the sound speed of the tensor modes is 1 , while the sound speed for scalar perturbations when dark matter is the dominant matter field is @xmath100 .
( note that this is not the case if the matter field is a scalar field in which case @xmath101 ; a small sound speed only arises if dark matter can be treated hydrodynamically . )
the differing sound speed for scalar and tensor modes has two important ramifications .
first , the amplitude of vacuum quantum fluctuations of short - wavelength perturbations @xmath102 ( with @xmath103 denoting either scalar or tensor perturbations ) , _ k ( ) = e^- i c_s k , depends on their sound speed .
in particular , a small sound speed leads to vacuum quantum fluctuations with a large amplitude
. therefore , vacuum quantum fluctuations for scalar perturbations will have a much larger amplitude than those for tensor perturbations .
this will remain true both inside the horizon while these modes oscillate freely , and also outside where this relative difference in amplitudes will be frozen once the modes exit their sound horizon . for this reason ,
the dark matter bounce scenario predicts a small tensor - to - scalar ratio .
second , for a given fourier wavenumber @xmath45 , the tensor mode will exit its sound horizon at a later time than the scalar mode with the same wavenumber .
this is because the sound horizon in cosmology is given by the sound speed of the perturbation divided by the hubble rate , @xmath104 .
a given fourier mode with wavenumber @xmath45 ( and physical wavelength @xmath105 ) exits its sound horizon when @xmath106 ; clearly , for a given @xmath45 the perturbations with the smallest sound speed will exit their horizon before other perturbations with a greater sound speed . as a consequence of this ,
the tensor modes exit their sound horizon at a later time than the scalar perturbations , at which time @xmath31 has increased .
if the tensor modes also exit their sound horizon when dark matter dominates the dynamics ( despite exiting at a later time than the scalar perturbations ) , then following the same calculation as for scalar perturbations it is easy to verify that the tensor spectral index is given by n_t = 12 ( _ t ) , where @xmath107 denotes the time that the tensor modes exited their horizon . in particular , if @xmath108 is negative then the tensor modes will also be scale - invariant with a slight red tilt .
furthermore , since @xmath31 increases as the space - time contracts , it follows that n_s - 1 n_t .
the exact difference between the two will depend on the detailed properties of the dark matter and dark energy in the model .
since the effective equation of state must evolve very slowly for models of this type to be viable , it seems likely that in these cases @xmath109 and @xmath110 will be very close ; however , this has to be checked on a case by case basis . also note that this is different from the consistency relation in single - field inflation , @xmath111 , and provides another possible mechanism for observations to differentiate between these two scenarios .
this may be tested in the near future by the high - precision measurement of polarization fluctuations in the cmb by forthcoming experiments , namely , the next generation of the bicep project , nasa s primordial inflation explorer ( pixie ) @xcite , and the ali project under design @xcite .
another possibility if @xmath34 is sufficiently small is that the tensor modes may exit their horizon at a much later time when the radiation field dominates the dynamics of the background space - time . in this case
, @xmath112 and the tensor perturbations will have a very blue spectrum , @xmath113 .
there are two main differences between matter bounce scenarios and inflation in terms of generating non - gaussianities .
first , the curvature perturbation will grow after it exits its horizon during matter contraction , rather than being conserved at super - horizon scales as it is in an inflationary background .
a consequence of the growing super - horizon mode leads to a different scaling of @xmath114 , and hence to a different shape of non - gaussianity .
second , again unlike in an inflationary background , the slow - roll parameter defined by @xmath115 is not a small quantity in matter bounce models .
rather , in a contracting matter - dominated space - time , @xmath116 .
for this reason , higher order terms of @xmath117 in the third order interaction lagrangian can not be neglected , as is the case in inflation , and their contribution to the bispectrum also leads to new features in the resulting primordial non - gaussianities .
see @xcite for further details . for the case
that the sound speed of the curvature perturbation is 1 , the above effects lead to @xmath118 amplitudes of non - gaussianities for the matter bounce scenario in three different limits @xcite : f_nl^local=- , f_nl^equil=- , f_nl^folded=- , which are consistent with the latest observational constraints @xcite .
these predictions are clearly distinct from the prediction of single - field slow - roll inflation , @xmath119 . in addition
, the shape of the primordial non - gaussianity is also quite unique in the matter bounce scenario , although the local form still dominates .
however , the above results were obtained for the case where the scalar perturbations are generated by a canonical scalar field with @xmath120 .
( note that in the matter bounce scenario with a canonical field , one can not get small enough @xmath114 and @xmath1 at the same time @xcite ; this is another justification to consider matter fields with a small sound speed , in additional to the obvious physical motivations . )
further work is required to determine how exactly these predictions will change for the case where the matter field is treated hydrodynamically with a small sound speed . at any rate ,
it is clear that non - gaussianities are another observable that may distinguish between matter bounce scenarios with dark matter and dark energy , and other early universe cosmologies like inflation .
there exist a number of cosmological observables that can constrain the parameter space of bouncing cosmologies , in addition to observational signatures from primordial fluctuations .
for example , magnetogenesis can be used to obtain a constraint on the relation of the energy scale of the bounce to the number of effective e - folds of the contracting phase @xcite , and it may also be possible to determine the energy scale of the bounce through dark matter searches @xcite .
as is well known , flrw space - times necessarily contain an initial big - bang singularity in an expanding universe ( or a final big - crunch singularity in a contracting universe ) if their dynamics are governed by the einstein equations and their matter fields satisfy the null energy condition @xcite .
therefore , a bouncing flrw space - time is only possible if either ( i ) the einstein equations are modified in some way , or ( ii ) at least one of the matter fields violates the null energy condition .
both of these possibilities have been studied in some detail in the literature , and a complete review of this field is beyond the scope of this paper . instead
, here we will briefly review some of the main proposals that lead to a bouncing cosmology : loop quantum cosmology , string cosmology , @xmath3 gravity , a ghost condensate scalar field , and the fermi bounce .
other theories that can also cause a bounce include hoava - lifshitz gravity @xcite , gauss - bonnet theories @xcite , modified teleparallel gravity theories @xcite , and lee - wick matter fields @xcite .
loop quantum cosmology ( lqc ) is motivated by loop quantum gravity ( lqg ) , a background independent non - perturbative theory of quantum gravity . in lqc ,
cosmological space - times like the spatially flat flrw space - time are quantized using the techniques of lqg . to be specific ,
there are two key ingredients : first , the fundamental variables of the theory are holonomies of the ashtekar - barbero connection @xmath121 and areas of surfaces , and second , the field strength operator is defined by taking the holonomy of @xmath121 around a loop of minimal area given by the smallest non - zero eigenvalue of the area operator in lqg @xcite . following this procedure
gives a well - defined quantum theory and makes it possible to explicitly calculate quantum gravity effects in this setting .
one of the most striking results is that the big - bang and big - crunch singularities are generically resolved and are replaced by a bounce .
furthermore , for wave functions that at some initial time are sharply peaked around a classical solution ( where the space - time curvature is small compared to the planck scale ) , their evolution is very simple : the spread of the wave function does not significantly grow , and the dynamics of the expectation value of the observables of interest ( the scale factor and the energy density of the matter field ) satisfy the effective friedmann equation [ lqc - fr ] h^2 = ( 1 - ) , where @xmath122 is the critical energy density that determines the energy scale ( or , equivalently , the curvature scale ) , at which the bounce occurs .
the continuity equation is not modified by quantum gravity effects .
lqc was developed to study homogeneous space - times , and determining the dynamics of cosmological perturbations in lqc is not immediate . in the case
that the perturbations of interest are long - wavelength fourier modes ( which is the case for the matter bounce scenario in lqc at the time that quantum gravity effects are important , i.e. , during the bounce ) , then the separate universe approximation @xcite can be used .
the idea is to divide the space - time into a large number of ( super - horizon - sized ) homogeneous patches , and then the long - wavelength cosmological perturbations are encoded in the differences in the homogeneous parameters describing each patch .
this is particularly useful for lqc , as now each patch can be quantized in the standard lqc fashion ( since each patch is homogeneous ) , and the equations of motion for the long - wavelength perturbations can be derived from the effective friedmann equations in each patch @xcite , giving v_k - v_k = 0 . although the form of the long - wavelength mukhanov - sasaki equation is not modified by lqc effects ,
it is important to keep in mind that the evolution of @xmath123 is determined by , not general relativity .
the evolution of scalar and tensor perturbations across the lqc bounce has been calculated in a number of contexts , and in particular in some realizations of the matter bounce scenario .
an interesting result is that , in addition to providing the required bounce for the scenario to be possible , lqc also tends to suppress the amplitude of the tensor perturbations with respect to the scalar perturbations @xcite .
this effect may allow future observations to probe the nature of the bounce in this scenario .
string theory also suggests that quantum gravity effects can resolve the big - bang singularity and replace it by a bounce .
this can perhaps most directly be seen by studying the thermal properties of a string gas on a space - time with a number of compactified dimensions using the euclidean description where the time dimension is compactified on a circle of radius @xmath124 ; to be specific here we shall consider weakly coupled @xmath125 superstrings compactified to 4 dimensions .
a careful analysis shows that the resulting one - loop partition function @xmath126 is always finite and satisfies the thermal duality ( t - duality ) @xcite z(r ) = z ( ) , where @xmath127 is of the order of the string length . at the critical point @xmath128 ,
some thermal string states become massless and form a condensate state . in cases where the temperature is time - dependent , a time slice with this condensate state will form a space - like brane ( s - brane ) . an important consequence of the t - duality is that there is a maximal temperature @xmath129 when @xmath130 , and the string gas is cold both for @xmath131 and @xmath132 .
these two limiting cases correspond respectively to a regime where the energy of the strings is mostly concentrated in their momentum and a regime where their energy is concentrated in windings around the compact dimensions .
therefore , it is possible for a process to occur where the string gas starts in the winding regime at a cold temperature , then the temperature is slowly increased until the maximal temperature @xmath129 is reached , at which point there is a stringy phase transition to the momentum regime and the temperature gradually decreases once more .
if such a process occurs in a cosmological space - time , the dynamics of the space - time can be determined from the thermodynamical properties of the string gas since the thermal entropy of the matter fields in a co - moving volume of a four - dimensional homogeneous space - time ( assuming the compactified dimensions are not dynamical ) , s = a^3 ~(at)^3 , is conserved . for this reason ,
@xmath133 must be constant , and since @xmath134 starts small , grows to its maximal value @xmath129 , and then decreases again , it follows that the scale factor is initially large , contracts to some minimal value ( where the s - brane appears ) , and then expands again ; giving a bouncing cosmology @xcite . at weak string coupling
, there exists an effective action for the cosmological dynamics which is valid in both low temperature regimes as well as close to @xmath129 @xcite .
this action can be perturbed to second order in perturbations around the flrw solution in order to calculate the evolution of cosmological perturbations from the pre - bounce contracting phase through the s - brane mediating the bounce to the expanding post - bounce phase @xcite . in particular
, this string cosmology can provide a stringy realization of the matter bounce scenario .
in addition , string theory has also been invoked in the context of ekpyrotic cosmologies in order to obtain a bounce to connect the big - crunch and big - bang singularies ; in this case the singularities are assumed to correspond to brane collisions in a higher - dimensional theory @xcite .
it may be possible to use this type of a cosmological bounce for the matter bounce scenario as well .
a simple family of modified gravity theories are the @xmath3 theories where the einstein - hilbert action is replaced by s = d^4x f(r ) .
the inclusion of higher order terms in the ricci scalar @xmath124 in the action is often motivated by two main observations : first , adding new terms in the space - time curvature could explain the observations typically associated to dark matter and/or dark energy , and second , since the einstein - hilbert action is not renormalizable , any consistent theory of quantum gravity is expected to contain higher order curvature terms in the action that become important near the planck scale @xcite . while @xmath3 theories are obviously not the most general such theory , they are among the simplest such modifications and
are often viewed as a good first step in understanding the effect of adding additional terms to the einstein - hilbert action .
the equations of motion of @xmath3 theories are @xcite
f r _ - g _ - [ _ _ - g _ ] f = 8 g t _ , where @xmath135 . the modifications in the einstein equations clearly generate modifications in the raychaudhuri equation , which allows the standard singularity theorems of general relativity to be avoided , and in particular , for some rather simple choices of @xmath3 , can generate a bounce @xcite .
in fact , it is even possible to use @xmath3 gravity to mimic very specific bouncing scenarios , including matter bounce scenarios .
this has been done explicitly , e.g. , for the lqc matter bounce scenario @xcite . at first
, it may appear that it will be difficult to determine the dynamics of cosmological perturbations in a general @xmath3 theory , but there exists a helpful short cut to do this . via a clever change of variables ,
= f g _ g _ , d = , @xmath3 theories can be rewritten in what is called the einstein frame , where the action is @xcite s = d^4x , where the ` potential ' is @xmath136 .
this rewriting of the action makes it clear that the usual equations of motion for cosmological perturbations from general relativity can be used , although with an additional ` scalar field ' @xmath137 @xcite . using this trick ,
it is possible to calculate how the pre - bounce form of the cosmological perturbations is affected by the bounce generated by some @xmath3 theory ( although to the best of our knowledge , this has not yet been done ) .
bounces can also be caused by matter fields that violate the null energy condition ; one such possibility is a ghost condensate scalar field @xmath103 @xcite , which is typically viewed as an effective field theory used to generate a bounce that is hoped to capture the key effects of the bounce , rather than as coming from a fundamental theory . in this effective field framework ,
an ( almost ) healthy version of non - singular bounce cosmology can be obtained by introducing a horndeski - type operator and a dynamical ghost condensate operator together .
typically , the lagrangian of this type of model may be expressed in the so - called kinetic gravity braiding ( kgb ) form @xcite as [ l_kgb ] l = k ( , x ) + g(x)+ l_4 + l_5 , where the specific forms of @xmath138 and @xmath139 are not important for our purposes here , and the operators @xmath140 and @xmath141 are chosen to be @xcite @xmath142 x + \f{\beta x^2}{m_{\rm pl}^4 } -v(\phi)~ , \\
g(x ) & = \f{\gamma x}{m_{\rm pl}^3 } ~.\end{aligned}\ ] ] here @xmath143 is the regular kinetic term for the scalar field , while @xmath144 and @xmath145 are real - valued parameters and @xmath146 is the usual dalembertian operator .
the phase of ghost condensation begins when the function @xmath147 becomes larger than 1 for a short time , and this can give rise to a non - singular bounce .
note that the @xmath140 term contains @xmath148 , which stabilizes the kinetic energy at high energy scales for @xmath149 .
it is possible to choose @xmath147 to be small far away from the bounce , in which case the lagrangian approaches the standard canonical form at low energy scales away from the bounce . around the same time as the ghost condensate phase occurs
, the cosmic time derivative of the scalar field @xmath150 typically reaches its maximal value near the bounce time and as a result the square of the sound speed parameter ( which contains a term @xmath151 ) decreases to a negative value at this time ; for a large class of parameter choices , near the bounce @xmath152 @xcite .
this phenomenon may at first appear problematic since a negative sound speed squared causes an exponential growth in the amplitude of short - wavelength cosmological perturbations , potentially leading to an instability in the theory .
( note that this does not affect the long - wavelength modes that are of observational interest .
) however , if this phase only exists for a short time , then the growth in the amplitude of the short - wavelength modes also only last for a short time and therefore remains under control if there is a minimal wavelength cutoff @xcite .
it is interesting to note that this phenomenon also appears to occur near the bounce in lqc as a result of quantum gravity effects ( although only for trans - planckian modes , if they exist ) .
it is possible to avoid the problem of a negative value for the sound speed squared in this effective framework by choosing other non - canonical kinetic operators @xcite , but there is necessarily a gradient instability , a ghost instability , or a singularity at some time ( although not necessarily during the bounce ) in bouncing cosmologies of this type @xcite .
nonetheless , the effective framework used for the ghost condensate remains consistent , even when these instabilities arise , due to the effective field theory cutoff @xcite .
interestingly , a comparison of realizations of the matter bounce scenario in lqc and in the effective ghost condensate framework showed that the dynamics of both the background space - time and the cosmological perturbations are quite similar in the two cases @xcite .
for example , the scale factors behave in similar fashions , and also the sound speed parameter becomes negative for a short time during both the lqc bounce as well as the ghost condensate bounce .
for this reason , it appears reasonable to expect that the effective field theory approach described here may indeed mimic the quantum gravity effects that arise at high energy scales .
fermi bounce models are based upon well known theories of particle physics that have been extensively tested on a flat gravitational background by means of high - energy terrestrial experiments .
their extension to curved space - time naturally allows for the possibility of violating the null energy condition and hence of causing a cosmological bounce .
one particularly interesting possible action for fermions coupled to gravity , known as the einstein - cartan - holst - sciama - kibble theory ( echsk ) , is given by the sum of the first order gravitational einstein - hilbert action , the holst topological term , and a non - minimally coupled covariant dirac action @xcite .
in this first order formalism of the theory , the spin - connection must have a torsionful part @xcite . here
, we will consider echsk models in the first order formalism for which the gravitational part of the action is _
holst = d^4x |e| e^_ie^_j p^ij_klf_^kl ( ) , where @xmath153 is the field - strength of the lorentz spin - connection @xmath154 , and @xmath155}_{l } - \epsilon^{ij}{}_{kl } / ( 2 \gamma)$ ] contains the barbero
immirzi parameter @xmath145 .
the dirac action is @xmath156 , with _
d = + h.c . , where @xmath157 is the non - minimal coupling parameter .
it is possible to integrate out the torsionful part of the spin - connection from the echsk action @xcite using the cartan equation , which can be recovered by varying the echsk action with respect to the spin - connection @xmath154 @xcite .
the result is a total action @xmath158 containing the einstein - hilbert action @xmath159 ( where @xmath160 is the field strength of the metric - compatible connection @xmath161^{ij}$ ] ) and the dirac action @xmath162 ( where @xmath163 denotes the covariant derivative with respect to @xmath161^{ij}$ ] ) .
the additional @xmath164 term encodes a four - fermion interaction potential , s_int = - 8 g d^4 x |e| j_5^l j_5^m _ lm , where @xmath165 is the axial current , and @xmath166\!/[16\,(\gamma^2 + 1)].$ ] importantly , @xmath167 can generate violations of the null energy condition is also interesting , as in a cosmological context the repulsive potential of the fermi liquids can sustain an accelerated phase of expansion of the universe @xcite . ] and can cause a bounce in cosmological space - times @xcite as well as , for some choices of parameters , ensure that black holes can not form @xcite . for suitable choices of fermion number density and bare mass , away from the bounce
the dynamics of the background space - time are the same as those of a matter - dominated flrw cosmology . in this case ,
scale - invariant cosmological perturbations are generated during the contracting pre - bounce phase , and are frozen after the bounce @xcite ( for a recent discussion on cosmological perturbation theory with fermionic fields , see @xcite ) .
it is also possible to include additional fermion fields , in which case the amplitude of the curvature perturbations is increased by a curvaton - like mechanism @xcite .
finally , note that while ghost condensates always include a gradient instability or ghost degrees of freedom @xcite , this does not appear to be the case for the fermi bounce .
indeed , since the fermionic action is only first - order in space - time derivatives , the fermi bounce has different properties than the ghost condensate bounces .
nonetheless , further work is still needed in order to verify that no instabilities arise in the fermi bounce .
the matter bounce scenario is an alternative to inflation where dark matter and dark energy can be included in a natural way . in particular , in the presence of dark matter and dark energy in a contracting universe , cosmological perturbations that exit their sound horizon when the effective equation of state is slightly negative will become nearly scale - invariant with a slight red tilt , as is observed in the cmb .
importantly , this cosmological scenario is falsifiable , as explained in sec .
[ s.obs ] , since it predicts a positive running of the scalar spectral index , tensor perturbations are found to have a redder spectrum than scalar perturbations , and there are new features in the predicted non - gaussianities .
the bounce can be generated by new physics at high energies , whether due to quantum gravity effects as in lqc and string cosmology , or particle physics effects as in the fermi bounce .
it is also possible to study the bounce at a more effective level with @xmath3 gravity or a ghost condensate scalar field .
an important question is whether it is possible to differentiate between the different types of bounces through observations .
one possibility in this direction is that lqc tends to suppress the tensor - to - scalar ratio during the bounce ; further work is needed in order to determine if any of the other types of bounces leave their own signatures in observables .
the predicted running of the scalar spectral index @xmath88 can be used to constrain different realizations of the matter bounce scenario , and in fact rules out the @xmath4cdm bounce scenario which predicts @xmath88 to be larger than what observations allow . a simple interacting dark energy model , reviewed in sec .
[ s.int ] , is viable since the interactions between dark matter and dark energy ensure that the effective equation of state evolves more slowly , and therefore predicts a smaller value for @xmath88 which is compatible with the latest observational constraints .
although this interacting dark energy model is phenomenologically interesting , the model is an effective one in which the dynamics of dark matter and dark energy are treated from a hydrodynamical perspective .
one important open problem is to go beyond this effective description of dark matter and dark energy . a natural way to overcome these shortcomings
is to consider the possibility that matter fields belonging to the standard model of particle physics could perhaps reproduce the required background and perturbation features , as suggested by @xcite . in particular
, it will be important to determine if these particle physics models of ( interacting ) dark matter and dark energy generate any distinct effects in the primordial curvature and tensor perturbations that could be tested by observations .
we would like to thank robert brandenberger , francis duplessis , jean - luc lehners and jerome quintin for valuable discussions .
the work of yfc and dgw is supported in part by the chinese national youth thousand talents program ( grant no .
kj2030220006 ) , by the ustc start - up funding ( grant no .
ky2030000049 ) , by the nsfc ( grant no .
11421303 ) , and by the fund for fostering talents in basic science of the nsfc ( grant no .
j1310021 ) .
am wishes to acknowledge support by the shanghai municipality , through the grant no .
kbh1512299 , and by fudan university , through the grant no .
dgw is also supported by a de sitter fellowship of the netherlands organization for scientific research ( nwo ) .
ewe is supported by a grant from the john templeton foundation .
part of the numerical computations are operated on the computer cluster linda in the particle cosmology group at ustc .
r. h. brandenberger , `` the matter bounce alternative to inflationary cosmology , '' http://arxiv.org/abs/1206.4196[arxiv:1206.4196 ] .
cai , e. mcdonough , f. duplessis , and r. h. brandenberger , `` two field matter bounce cosmology , '' jcap * 1310 * ( 2013 ) 024 , http://arxiv.org/abs/1305.5259[arxiv:1305.5259 ] . , p. ade _ et al .
_ , `` planck 2015 results
cosmological parameters , '' http://arxiv.org/abs/1502.01589[arxiv:1502.01589 ] .
, p. ade _ et al .
_ , `` joint analysis of bicep2/@xmath168 and @xmath169 data , '' phys .
* 114 * ( 2015 ) 101301 , http://arxiv.org/abs/1502.00612[arxiv:1502.00612 ] .
j. quintin , z. sherkatghanad , y .- f .
cai , and r. h. brandenberger , `` evolution of cosmological perturbations and the production of non - gaussianities through a nonsingular bounce : indications for a no - go theorem in single field matter bounce cosmologies , '' phys .
* d92 * ( 2015 ) 063532 , http://arxiv.org/abs/1508.04141[arxiv:1508.04141 ] .
e. wilson - ewing , `` the matter bounce scenario in loop quantum cosmology , '' jcap * 1303 * ( 2013 ) 026 , http://arxiv.org/abs/1211.6269[arxiv:1211.6269 ] .
f . cai and e. wilson - ewing , `` a @xmath4cdm bounce scenario , '' jcap * 1503 * ( 2015 ) 006 , http://arxiv.org/abs/1412.2914[arxiv:1412.2914 ] .
cai , f. duplessis , d. a. easson , and d .-
wang , `` searching for a matter bounce cosmology with low redshift observations , '' phys .
* d93 * ( 2016 ) 043546 , http://arxiv.org/abs/1512.08979[arxiv:1512.08979 ] .
cai , r. brandenberger , and x. zhang , `` the matter bounce curvaton scenario , '' jcap * 1103 * ( 2011 ) 003 , http://arxiv.org/abs/1101.0822[arxiv:1101.0822 ] .
a. fertig , j .- l .
lehners , e. mallwitz , and e. wilson - ewing , `` converting entropy to curvature perturbations after a cosmic bounce , '' http://arxiv.org/abs/1607.05663[arxiv:1607.05663 ] .
e. wilson - ewing , `` separate universes in loop quantum cosmology : framework and applications , '' int .
j. mod .
* d25 * ( 2016 ) 1642002 , http://arxiv.org/abs/1512.05743[arxiv:1512.05743 ] .
a. balbi , m. bruni , and c. quercellini , `` lambda - alpha dm : observational constraints on unified dark matter with constant speed of sound , '' phys .
* d76 * ( 2007 ) 103519 , http://arxiv.org/abs/astro-ph/0702423[arxiv:astro-ph/0702423 ] .
p. p. avelino and v. m. c. ferreira , `` constraints on the dark matter sound speed from galactic scales : the cases of the modified and extended chaplygin gas , '' phys
* d91 * ( 2015 ) 083508 , http://arxiv.org/abs/1502.07583[arxiv:1502.07583 ] .
j. de haro and y .- f .
cai , `` an extended matter bounce scenario : current status and challenges , '' gen .
rel . grav .
* 47 * ( 2015 ) 95 , http://arxiv.org/abs/1502.03230[arxiv:1502.03230 ] .
s. d. odintsov and v. k. oikonomou , `` @xmath4cdm bounce cosmology without @xmath4cdm : the case of modified gravity , '' phys . rev . *
d91 * ( 2015 ) 064036 , http://arxiv.org/abs/1502.06125[arxiv:1502.06125 ] .
lehners and e. wilson - ewing , `` running of the scalar spectral index in bouncing cosmologies , '' jcap * 1510 * ( 2015 ) 038 , http://arxiv.org/abs/1507.08112[arxiv:1507.08112 ] .
p. c. ferreira and d. pavon , `` thermodynamics of nonsingular bouncing universes , '' eur .
j. * c76 * ( 2016 ) 37 , http://arxiv.org/abs/1509.03725[arxiv:1509.03725 ] .
r. h. brandenberger , y .- f .
cai , s. r. das , e. g. m. ferreira , i. a. morrison , and y. wang , `` fluctuations in a cosmology with a space - like singularity and their gauge theory dual description , '' http://arxiv.org/abs/1601.00231[arxiv:1601.00231 ] .
s. nojiri , s. d. odintsov , and v. k. oikonomou , `` bounce universe history from unimodular @xmath170 gravity , '' phys . rev .
* d93 * ( 2016 ) 084050 , http://arxiv.org/abs/1601.04112[arxiv:1601.04112 ] .
s. d. odintsov and v. k. oikonomou , `` deformed matter bounce with dark energy epoch , '' http://arxiv.org/abs/1606.03689[arxiv:1606.03689 ] .
v. f. mukhanov , h. feldman , and r. h. brandenberger , `` theory of cosmological perturbations , '' phys .
* 215 * ( 1992 ) 203333 .
v. mukhanov , _ physical foundations of cosmology_. cambridge university press , cambridge , 2005 . y .-
cai , r. brandenberger , and x. zhang , `` preheating a bouncing universe , '' phys .
* b703 * ( 2011 ) 2533 , http://arxiv.org/abs/1105.4286[arxiv:1105.4286 ] .
j. quintin , y .- f .
cai , and r. h. brandenberger , `` matter creation in a nonsingular bouncing cosmology , '' phys .
* d90 * ( 2014 ) 063507 , http://arxiv.org/abs/1406.6049[arxiv:1406.6049 ] .
e. j. copeland , m. sami , and s. tsujikawa , `` dynamics of dark energy , '' int . j. mod .
d15 * ( 2006 ) 17531936 , http://arxiv.org/abs/hep-th/0603057[arxiv:hep-th/0603057 ] .
cai , e. n. saridakis , m. r. setare , and j .- q .
xia , `` quintom cosmology : theoretical implications and observations , '' phys .
* 493 * ( 2010 ) 160 , http://arxiv.org/abs/0909.2776[arxiv:0909.2776 ] .
m. li , x .- d .
li , s. wang , and y. wang , `` dark energy , '' commun .
* 56 * ( 2011 ) 525604 , http://arxiv.org/abs/1103.5870[arxiv:1103.5870 ] .
l. p. chimento , a. s. jakubi , d. pavon , and w. zimdahl , `` interacting quintessence solution to the coincidence problem , '' phys . rev .
* d67 * ( 2003 ) 083513 , http://arxiv.org/abs/astro-ph/0303145[arxiv:astro-ph/0303145 ] .
l. amendola , `` coupled quintessence , '' phys .
* d62 * ( 2000 ) 043511 , http://arxiv.org/abs/astro-ph/9908023[arxiv:astro-ph/9908023 ] .
d. comelli , m. pietroni , and a. riotto , `` dark energy and dark matter , '' phys . lett .
* b571 * ( 2003 ) 115120 , http://arxiv.org/abs/hep-ph/0302080[arxiv:hep-ph/0302080 ] .
x. zhang , `` coupled quintessence in a power - law case and the cosmic coincidence problem , '' mod .
* a20 * ( 2005 ) 2575 , http://arxiv.org/abs/astro-ph/0503072[arxiv:astro-ph/0503072 ] .
cai and a. wang , `` cosmology with interaction between phantom dark energy and dark matter and the coincidence problem , '' jcap * 0503 * ( 2005 ) 002 , http://arxiv.org/abs/hep-th/0411025[arxiv:hep-th/0411025 ] .
k . guo , r .-
cai , and y .- z .
zhang , `` cosmological evolution of interacting phantom energy with dark matter , '' jcap * 0505 * ( 2005 ) 002 , http://arxiv.org/abs/astro-ph/0412624[arxiv:astro-ph/0412624 ] .
, t. delubac _ et al .
_ , `` baryon acoustic oscillations in the ly forest of boss dr11 quasars , '' astron . astrophys .
* 574 * ( 2015 ) a59 , http://arxiv.org/abs/1404.1801[arxiv:1404.1801 ] .
v. salvatelli , n. said , m. bruni , a. melchiorri , and d. wands , `` indications of a late - time interaction in the dark sector , '' phys .
lett . * 113 * ( 2014 ) 181301 , http://arxiv.org/abs/1406.7297[arxiv:1406.7297 ] .
e. abdalla , e. g. m. ferreira , j. quintin , and b. wang , `` new evidence for interacting dark energy from boss , '' http://arxiv.org/abs/1412.2777[arxiv:1412.2777 ] .
j. vliviita and e. palmgren , `` distinguishing interacting dark energy from wcdm with cmb , lensing , and baryon acoustic oscillation data , '' jcap * 1507 * ( 2015 ) 015 , http://arxiv.org/abs/1504.02464[arxiv:1504.02464 ] .
p. don , a. marcian , y. zhang , and c. antolini , `` yang - mills condensate as dark energy : a nonperturbative approach , '' phys . rev .
* d93 * ( 2016 ) 043012 , http://arxiv.org/abs/1509.05824[arxiv:1509.05824 ] .
a. addazi , p. don , and a. marcian , `` dark energy and dark matter from yang - mills condensate and the peccei - quinn mechanism , '' http://arxiv.org/abs/1602.01772[arxiv:1602.01772 ] .
s. alexander , a. marcian , and z. yang , `` invisible qcd as dark energy , '' http://arxiv.org/abs/1602.06557[arxiv:1602.06557 ] .
a. addazi , a. marciano , and s. alexander , `` a unified picture of dark matter and dark energy from invisible qcd , '' http://arxiv.org/abs/1603.01853[arxiv:1603.01853 ] .
j. valiviita , e. majerotto , and r. maartens , `` instability in interacting dark energy and dark matter fluids , '' jcap * 0807 * ( 2008 ) 020 , http://arxiv.org/abs/0804.0232[arxiv:0804.0232 ] .
he , b. wang , and e. abdalla , `` stability of the curvature perturbation in dark sectors mutual interacting models , '' phys .
* b671 * ( 2009 ) 139145 , http://arxiv.org/abs/0807.3471[arxiv:0807.3471 ] .
a. a. costa , x .- d .
xu , b. wang , e. g. m. ferreira , and e. abdalla , `` testing the interaction between dark energy and dark matter with planck data , '' phys . rev .
* d89 * ( 2014 ) 103531 , http://arxiv.org/abs/1311.7380[arxiv:1311.7380 ] .
g. cabass , e. di valentino , a. melchiorri , e. pajer , and j. silk , `` running the running , '' http://arxiv.org/abs/1605.00209[arxiv:1605.00209 ] .
cai , `` exploring bouncing cosmologies with cosmological surveys , '' sci .
china phys .
* 57 * ( 2014 ) 14141430 , http://arxiv.org/abs/1405.1369[arxiv:1405.1369 ] .
t. kobayashi and f. takahashi , `` running spectral index from inflation with modulations , '' jcap * 1101 * ( 2011 ) 026 , http://arxiv.org/abs/1011.3988[arxiv:1011.3988 ] .
a. kogut _ et al .
_ , `` the primordial inflation explorer ( pixie ) : a nulling polarimeter for cosmic microwave background observations , '' jcap * 1107 * ( 2011 ) 025 , http://arxiv.org/abs/1105.2044[arxiv:1105.2044 ] .
f . cai and x. zhang , `` probing the origin of our universe through primordial gravitational waves by ali cmb project , '' sci .
china phys .
* 59 * ( 2016 ) 670431 , http://arxiv.org/abs/1605.01840[arxiv:1605.01840 ] .
cai , w. xue , r. brandenberger , and x. zhang , `` non - gaussianity in a matter bounce , '' jcap * 0905 * ( 2009 ) 011 , http://arxiv.org/abs/0903.0631[arxiv:0903.0631 ] . , p. a. r. ade _ et al .
_ , `` planck 2015 results .
constraints on primordial non - gaussianity , '' http://arxiv.org/abs/1502.01592[arxiv:1502.01592 ] .
qian , y .- f .
cai , d. a. easson , and z .- k .
guo , `` magnetogenesis in bouncing cosmology , '' http://arxiv.org/abs/1607.06578[arxiv:1607.06578 ] . c. li , r. h. brandenberger , and y .- k .
e. cheung , `` big - bounce genesis , '' phys
* d90 * ( 2014 ) 123535 , http://arxiv.org/abs/1403.5625[arxiv:1403.5625 ] .
e. cheung , j. u. kang , and c. li , `` dark matter in a bouncing universe , '' jcap * 1411 * ( 2014 ) 001 , http://arxiv.org/abs/1408.4387[arxiv:1408.4387 ] .
e. cheung and j. d. vergados , `` direct dark matter searches - test of the big bounce cosmology , '' jcap * 1502 * ( 2015 ) 014 , http://arxiv.org/abs/1410.5710[arxiv:1410.5710 ] . s. w. hawking and g. f. r. ellis , _ the large scale structure of space - time_. cambridge monographs on mathematical physics .
cambridge university press , 2011 .
r. brandenberger , `` matter bounce in horava - lifshitz cosmology , '' phys .
* d80 * ( 2009 ) 043516 , http://arxiv.org/abs/0904.2835[arxiv:0904.2835 ] .
f . cai and e. n. saridakis , `` non - singular cosmology in a model of non - relativistic gravity , '' jcap * 0910 * ( 2009 ) 020 , http://arxiv.org/abs/0906.1789[arxiv:0906.1789 ] .
k. bamba , a. n. makarenko , a. n. myagky , and s. d. odintsov , `` bouncing cosmology in modified gauss - bonnet gravity , '' phys .
* b732 * ( 2014 ) 349355 , http://arxiv.org/abs/1403.3242[arxiv:1403.3242 ] .
cai , s .- h .
chen , j. b. dent , s. dutta , and e. n. saridakis , `` matter bounce cosmology with the f(t ) gravity , '' class .
* 28 * ( 2011 ) 215011 , http://arxiv.org/abs/1104.4349[arxiv:1104.4349 ] .
cai , s. capozziello , m. de laurentis , and e. n. saridakis , `` f(t ) teleparallel gravity and cosmology , '' http://arxiv.org/abs/1511.07586[arxiv:1511.07586 ] .
f . cai , t .-
qiu , r. brandenberger , and x .- m .
zhang , `` a nonsingular cosmology with a scale - invariant spectrum of cosmological perturbations from lee - wick theory , '' phys .
* d80 * ( 2009 ) 023511 , http://arxiv.org/abs/0810.4677[arxiv:0810.4677 ] .
a. ashtekar , t. pawowski , and p. singh , `` quantum nature of the big bang : improved dynamics , '' phys .
* d74 * ( 2006 ) 084003 , http://arxiv.org/abs/gr-qc/0607039[arxiv:gr-qc/0607039 ] .
d. wands , k. a. malik , d. h. lyth , and a. r. liddle , `` a new approach to the evolution of cosmological perturbations on large scales , '' phys .
rev . * d62 * ( 2000 ) 043527 , http://arxiv.org/abs/astro-ph/0003278[arxiv:astro-ph/0003278 ] .
e. wilson - ewing , `` lattice loop quantum cosmology : scalar perturbations , '' class .
* 29 * ( 2012 ) 215013 , http://arxiv.org/abs/1205.3370[arxiv:1205.3370 ] . c. kounnas , h. partouche , and n. toumbas , `` thermal duality and non - singular cosmology in d - dimensional superstrings , '' nucl .
b855 * ( 2012 ) 280307 , http://arxiv.org/abs/1106.0946[arxiv:1106.0946 ] . c. kounnas , h. partouche , and n. toumbas , `` s - brane to thermal non - singular string cosmology , '' class .
* 29 * ( 2012 ) 095014 , http://arxiv.org/abs/1111.5816[arxiv:1111.5816 ] . c. kounnas and n. toumbas ,
`` aspects of string cosmology , '' pos * corfu2012 * ( 2013 ) 083 , http://arxiv.org/abs/1305.2809[arxiv:1305.2809 ] .
r. h. brandenberger , c. kounnas , h. partouche , s. p. patil , and n. toumbas , `` cosmological perturbations across an s - brane , '' jcap * 1403 * ( 2014 ) 015 , http://arxiv.org/abs/1312.2524[arxiv:1312.2524 ] .
j. khoury , b. a. ovrut , n. seiberg , p. j. steinhardt , and n. turok , `` from big crunch to big bang , '' phys .
* d65 * ( 2002 ) 086007 , http://arxiv.org/abs/hep-th/0108187[arxiv:hep-th/0108187 ] .
lehners , p. mcfadden , and n. turok , `` colliding branes in heterotic m - theory , '' phys .
rev . * d75 * ( 2007 ) 103510 , http://arxiv.org/abs/hep-th/0611259[arxiv:hep-th/0611259 ] . t. p. sotiriou and v. faraoni , `` f(r ) theories of gravity , '' rev . mod .
* 82 * ( 2010 ) 451497 , http://arxiv.org/abs/0805.1726[arxiv:0805.1726 ] .
s. carloni , p. k. s. dunsby , and d. m. solomons , `` bounce conditions in f(r ) cosmologies , '' class .
* 23 * ( 2006 ) 19131922 , http://arxiv.org/abs/gr-qc/0510130[arxiv:gr-qc/0510130 ] . c. barragan , g. j. olmo , and h. sanchis - alepuz , `` bouncing cosmologies in palatini f(r ) gravity , '' phys .
* d80 * ( 2009 ) 024016 , http://arxiv.org/abs/0907.0318[arxiv:0907.0318 ] .
k. bamba , a. n. makarenko , a. n. myagky , s. nojiri , and s. d. odintsov , `` bounce cosmology from @xmath170 gravity and @xmath170 bigravity , '' jcap * 1401 * ( 2014 ) 008 , http://arxiv.org/abs/1309.3748[arxiv:1309.3748 ] .
s. odintsov and v. oikonomou , `` matter bounce loop quantum cosmology from @xmath170 gravity , '' http://arxiv.org/abs/1410.8183[arxiv:1410.8183 ] . c. lin , r. h. brandenberger , and l. perreault levasseur , `` a matter bounce by means of ghost condensation , '' jcap * 1104 * ( 2011 ) 019 , http://arxiv.org/abs/1007.2654[arxiv:1007.2654 ] .
cai , d. a. easson , and r. brandenberger , `` towards a nonsingular bouncing cosmology , '' jcap * 1208 * ( 2012 ) 020 , http://arxiv.org/abs/1206.2382[arxiv:1206.2382 ] .
cai , r. brandenberger , and p. peter , `` anisotropy in a nonsingular bounce , '' class .
. grav . * 30 * ( 2013 ) 075019 , http://arxiv.org/abs/1301.4703[arxiv:1301.4703 ] .
m. koehn , j .- l .
lehners , and b. a. ovrut , `` cosmological super - bounce , '' phys . rev .
* d90 * ( 2014 ) 025005 , http://arxiv.org/abs/1310.7577[arxiv:1310.7577 ] . c. deffayet , o. pujolas , i. sawicki , and a. vikman , `` imperfect dark energy from kinetic gravity braiding , '' jcap * 1010 * ( 2010 ) 026 , http://arxiv.org/abs/1008.0048[arxiv:1008.0048 ] . l. battarra , m. koehn , j .- l .
lehners , and b. a. ovrut , `` cosmological perturbations through a non - singular ghost - condensate / galileon bounce , '' jcap * 1407 * ( 2014 ) 007 , http://arxiv.org/abs/1404.5067[arxiv:1404.5067 ] .
t. qiu , j. evslin , y .- f .
cai , m. li , and x. zhang , `` bouncing galileon cosmologies , '' jcap * 1110 * ( 2011 ) 036 , http://arxiv.org/abs/1108.0593[arxiv:1108.0593 ] .
d. a. easson , i. sawicki , and a. vikman , `` g - bounce , '' jcap * 1111 * ( 2011 ) 021 , http://arxiv.org/abs/1109.1047[arxiv:1109.1047 ] .
a. ijjas and p. j. steinhardt , `` classically stable non - singular cosmological bounces , '' http://arxiv.org/abs/1606.08880[arxiv:1606.08880 ] .
s. dubovsky , t. gregoire , a. nicolis , and r. rattazzi , `` null energy condition and superluminal propagation , '' jhep * 03 * ( 2006 ) 025 , http://arxiv.org/abs/hep-th/0512260[arxiv:hep-th/0512260 ] .
m. libanov , s. mironov , and v. rubakov , `` generalized galileons : instabilities of bouncing and genesis cosmologies and modified genesis , '' jcap * 1608 * ( 2016 ) 037 , http://arxiv.org/abs/1605.05992[arxiv:1605.05992 ]
. t. kobayashi , `` generic instabilities of nonsingular cosmologies in horndeski theory : a no - go theorem , '' phys .
* d94 * ( 2016 ) 043511 , http://arxiv.org/abs/1606.05831[arxiv:1606.05831 ] .
m. koehn , j .- l .
lehners , and b. ovrut , `` nonsingular bouncing cosmology : consistency of the effective description , '' phys .
* d93 * ( 2016 ) 103501 , http://arxiv.org/abs/1512.03807[arxiv:1512.03807 ] .
f . cai and e. wilson - ewing , `` non - singular bounce scenarios in loop quantum cosmology and the effective field description , '' jcap * 1403 * ( 2014 ) 026 , http://arxiv.org/abs/1402.3009[arxiv:1402.3009 ] .
i. l. shapiro , `` physical aspects of the space - time torsion , '' phys .
( 2002 ) 113 , http://arxiv.org/abs/hep-th/0103093[arxiv:hep-th/0103093 ] .
r. t. hammond , `` torsion gravity , '' rept .
* 65 * ( 2002 ) 599649 .
a. trautman , `` einstein - cartan theory , '' in _ encyclopedia of mathematical physics _
franoise , g. l. naber , and s. t. tsou , eds . , vol . 2 , pp .
elsevier , oxford , 2006 .
http://arxiv.org/abs/gr-qc/0606062[arxiv:gr-qc/0606062 ] .
a. perez and c. rovelli , `` physical effects of the immirzi parameter , '' phys
* d73 * ( 2006 ) 044013 , http://arxiv.org/abs/gr-qc/0505081[arxiv:gr-qc/0505081 ] .
l. freidel , d. minic , and t. takeuchi , `` quantum gravity , torsion , parity violation and all that , '' phys
* d72 * ( 2005 ) 104002 , http://arxiv.org/abs/hep-th/0507253[arxiv:hep-th/0507253 ] . c. bambi , d. malafarina , a. marcian , and l. modesto , `` singularity avoidance in classical gravity from four - fermion interaction , '' phys
b734 * ( 2014 ) 2730 , http://arxiv.org/abs/1402.5719[arxiv:1402.5719 ] .
s. alexander , t. biswas , and g. calcagni , `` cosmological bardeen - cooper - schrieffer condensate as dark energy , '' phys . rev .
* d81 * ( 2010 ) 043511 , http://arxiv.org/abs/0906.5161[arxiv:0906.5161 ] .
[ erratum : phys .
rev.d81,069902(2010 ) ] . s. alexander and t. biswas , `` the cosmological bcs mechanism and the big bang singularity , '' phys .
* d80 * ( 2009 ) 023501 , http://arxiv.org/abs/0807.4468[arxiv:0807.4468 ] .
n. j. popawski , `` nonsingular , big - bounce cosmology from spinor - torsion coupling , '' phys .
* d85 * ( 2012 ) 107502 , http://arxiv.org/abs/1111.4595[arxiv:1111.4595 ] .
s. alexander , c. bambi , a. marcian , and l. modesto , `` fermi - bounce cosmology and scale invariant power - spectrum , '' phys . rev .
* d90 * ( 2014 ) 123510 , http://arxiv.org/abs/1402.5880[arxiv:1402.5880 ] .
s. alexander , y .- f .
cai , and a. marcian , `` fermi - bounce cosmology and the fermion curvaton mechanism , '' http://arxiv.org/abs/1406.1456[arxiv:1406.1456 ] .
p. don and a. marcian , `` a note on the semiclassicality of cosmological perturbations , '' http://arxiv.org/abs/1605.09337[arxiv:1605.09337 ] . | we review matter bounce scenarios where the matter content is dark matter and dark energy .
these cosmologies predict a nearly scale - invariant power spectrum with a slightly red tilt for scalar perturbations and a small tensor - to - scalar ratio .
importantly , these models predict a positive running of the scalar index , contrary to the predictions of the simplest inflationary and ekpyrotic models , and hence could potentially be falsified by future observations .
we also review how bouncing cosmological space - times can arise in theories where either the einstein equations are modified or where matter fields that violate the null energy condition are included . |
factorization properties of qcd allow us to calculate exclusive processes provided a hard scale is available , either the three mandelstam variables @xmath0 are large as compared to a typical hadronic scale @xmath1 where @xmath2 is of order @xmath3 ( wide - angle processes ) or there is a highly virtual photon involved ( deeply virtual processes ) . in the space - like region typical hard processes are real and virtual compton scattering or photo- and electroproduction of mesons . in these cases
the process amplitudes factorize in a hard partonic subprocess ( e.g. @xmath4 ) and in soft hadronic matrix elements parameterized as generalized parton distributions ( gpds ) . for time - like processes , e.g. two - photon annihilations into pairs of hadrons
, an analogous factorization scheme holds @xcite . in this case
the soft hadronic matrix elements are so - called two - hadron distribution amplitudes , time - like versions of gpds , see fig .
[ fig:1 ] .
one may consider more complicated topologies than shown in fig .
[ fig:1 ] in which @xmath5 partons are emitted and reabsorbed from the hadrons . in order to have quasi - on - shell partons entering the hadrons , a compelling requirement for factorization , @xmath6
hard gluons are needed to be exchanged between the @xmath5 active partons .
for very large @xmath7 and @xmath8 one may treat the hadrons in valence quark approximation .
considering for instance @xmath9 with an active quark - antiquark pair ( @xmath10 ) . in this case
there is no spectator left and , hence , the soft physics is encoded in two meson distribution amplitudes instead of a time - like gpd .
this is an example of the so - called erbl factorization scheme which has been invented for @xmath11 in @xcite .
the arguments for factorization of @xmath9 in the wide - angle region have been given in @xcite .
for the ease of legibility let @xmath12 be charged pion ; the generalization to other pseudoscalar mesons is straightforward .
it is of advantage to work in a symmetric ( c.m . )
frame in which the pions move along the 1-axis .
the light - cone plus components of the meson momenta , @xmath13 and @xmath14 , are then equal : @xmath15 and the skewness , defined by @xmath16 , is 1/2 .
the momentum fraction of the active quark is as usual defined by @xmath17 ; the momenta fraction of the antiquark is @xmath18 . in order to achieve the factorization in a hard process @xmath19 and a soft transition @xmath20 ,
two assumptions are made in @xcite : i ) restricted transverse momenta , @xmath21 , for the active partons as well for the spectators , and ii ) all virtualities at the parton - hadron vertices are soft of order @xmath1 . with the help of these assumptions
one can show that the following requirements must hold : @xmath22 where @xmath23 describes the orientation of the active parton momenta in the 12 plane .
the second requirement has two solutions @xmath24 one can then derive a factorization formula for the helicity amplitudes @xmath25 where due to charge conjugation invariance @xmath26 the soft matrix element @xmath27 is expected to be strongly peaked when ( [ eq : requirements ] ) is fulfilled .
the two regions @xmath28 and @xmath29 where this is the case are related through a rotation by @xmath30 about the 3-axis of our coordinate frame .
the hard scattering kernel @xmath31 can be taylor expanded around the 1-axis and @xmath32 @xmath33 where the labels denote the helicities of the photons ( @xmath34 ) . in terms of the c.m
. scattering angle , @xmath35 , the first term in ( [ eq : exp ] ) is proportional to @xmath36 while the second one is @xmath37 .
it can then be shown @xcite that for the @xmath38 channel the first term in this expansion vanishes due to a conspiracy of charge conjugation and rotation by @xmath39 .
keeping therefore only the second term in ( [ eq : exp ] ) and perform the integrals in ( [ eq : ansatz ] ) one arrives at @xmath40 where the annihilation form factors encoding the soft physics , is defined by @xmath41 here , @xmath42 is the two - pion distribution amplitude @xcite defined as the fourier transform of a bilocal vacuum - two - pion matrix element of quark field operators .
this distribution amplitude also determines the electromagnetic form factor of the pion in the time - like region @xmath43 the energy dependences of the annihilation and electromagnetic form factors are not predicted in the handbag approach .
although the @xmath44 term in ( [ eq : exp ] ) is of the same order as the parton - off - shell effects we remain with the on - shell approximation .
therefore the above sketched approach to @xmath45 is to be considered as a model . the suppression of the leading term in ( [ eq : exp ] ) is special to @xmath46 ; it is due to a conspiracy of charge conjugation invariance and a rotation .
this conspiracy does not occur in other reactions and the leading term in ( [ eq : exp ] ) dominates .
for instance , in two - photon annihilations into baryon - antibaryon ( @xmath47 ) pairs @xcite the region @xmath48 corresponds to antiquark hadronization into a baryon which requires sea quarks with very high momentum fractions .
it is unlikely that such sea quarks exist .
another example is set by real compton scattering in the space - like region @xcite .
the regions @xmath49 ( @xmath50 ) and @xmath51 ( @xmath52 ) are related by charge conjugation but not by a rotation .
the above handbag result for the @xmath53 channel can easily be generalized to the production of other pairs of pseudoscalar mesons .
the corresponding differential cross section reads @xmath54 there are six pseudoscalar meson channels available in two - photon annihilations including altogether 18 annihilation form factors . to fix these form factors from experiment is a boring program .
however , due to flavor symmetry ( the meson pair couples to an @xmath55-spin singlet ) and due to the absence of isospin 2 ( and @xmath56-spin 2 ) states ( because the two photons annihilate via a quark - antiquark intermediate state ) there are only two independent annihilation form factors , say @xmath57 and @xmath58 , or , in other words , a valence and a non - valence form factor .
the combinations of the individual flavor contributions appearing in the various channels read in terms of the two independent form factors @xmath59 in addition there are a number of relations and triangular inequalities among the cross sections .
the belle collaboration has measured the wide - angle cross sections for the six pseudoscalar meson channels up to fairly large energies @xcite .
the data are compatible with a @xmath60 behavior for @xmath61 .
given that fact one has only to deal with the form factors or the respective integrated cross sections .
it turns out that the inequalities and relations following from flavor symmetry and isospin selection rules are satisfied within errors in general . from ( [ eq : symmetry_relations ] ) and from the statistical factor to be applied to the integrated cross section for identical particles it follows that the ratio of the @xmath62 and @xmath38 is predicted to be @xmath63 in the handbag approach . the data shown in fig.[fig : ratios ] clearly deviate from @xmath63 at low @xmath64 but seem to approach this value for @xmath65 within rapidly growing errors . a small @xmath66 admixture which may vanish for @xmath67 ,
can easily explain the deviation from @xmath63 .
assuming for example that the @xmath66 and @xmath68 amplitudes are in phase then an admixture @xmath69 is sufficient to explain the observed deviation .
thus , the dominance of the @xmath68 amplitude is fully consistent with experiment for large @xmath64 . in fig .
[ fig : ratios ] the ratio of the @xmath53 combination @xmath70 and the @xmath71 cross sections are shown and can be seen to be compatible with 1 according to the handbag approach , see ( [ eq : symmetry_relations ] ) .
note that in the above combination of @xmath53 cross sections the interference term between the @xmath66 and @xmath68 cancels .
for a few values of the energy the two basic form factors as well as their relative phase , @xmath72 , can directly be extracted from the data on , say , @xmath71 , @xmath73 and @xmath74 production . alternatively , in order to reduce errors , one may perform a combined energy - dependent fit .
this fit provides @xmath75 \label{eq : parameter}\end{aligned}\ ] ] ( @xmath76 ) .
the energy dependencies of the other channels ( @xmath38 , @xmath62 , @xmath77 ) are then fixed in agreement with experiment .
the relative phase varies rapidly in the charmonium region and approaches @xmath30 for @xmath67 . in concord with expectation , the non - valence form factor , @xmath78 ,
is much smaller than the valence one and decreases more rapidly with increasing energy than the other one .
hence , for @xmath67 only the valence form factor remains and the six cross sections all exhibit the same energy dependence and only differ in magnitude by charge factors .
for the range of energy measured by belle the energy dependence of the various cross sections differ quite a bit due to the different superpositions of the valence and non - valence form factors .
charged pions and kaons have the smallest contributions from the non - valence form factor .
therefore , their integrated cross sections drop only slightly faster then @xmath79 . on the other hand
, the @xmath73 ( @xmath80 ) cross section falls off rapidly with energy because of the relative strong contribution from the non - valence form factor in this case .
this is in agreement with experiment as can be seen from tab .
[ tab:1 ] where the effective powers of @xmath64 obtained by belle from fits @xmath81 to its data are shown .
[ cols="^,^,^,^,^,^ " , ] the process of interest has also been investigated in the erbl factorization scheme @xcite . in this approach
the mesons are treated in valence quark approximation and the amplitudes are given by a convolution of the distribution amplitudes for the mesons and a hard scattering kernel @xmath82 where @xmath83 is the decay constant of meson @xmath84 . as for the handbag
approach the cross section behaves as @xmath60 but its energy dependence is predicted to be @xmath85 , cf . tab .
[ tab:1 ] .
while this is in rough agreement with experiment for charged mesons the magnitude of the corresponding cross section is underestimated when it is evaluated from a distribution amplitude that is close to the asymptotic form , @xmath86 which is favored by current phenomenology and lattice results @xcite .
the production of neutral meson pairs is generically suppressed , since at leading order of @xmath87 the bulk of the amplitude is proportional to a charge factor @xmath88 for a meson with quark content @xmath89 @xcite .
explicit calculations @xcite yield values below 0.05 for the ratio of @xmath62 and @xmath38 cross sections which is significantly below the experimental results shown in fig .
[ fig : ratios ] .
because of charge factors the ratio of @xmath90 and @xmath71 cross sections is predicted to be even smaller . despite this suppression of the neutral meson pairs the energy dependencies of the corresponding cross sections is @xmath91 which is also in conflict with experiment , see tab .
[ tab:1 ] . for these reasons
it was suggested in @xcite that at belle energies the production of neutral meson pairs is dominated by a contribution other then the the erbl mechanism but a quantitative study of the new contribution is lacking .
the meson case can be straightforwardly generalized to the production of baryon - antibaryon pairs .
an important difference is that the cross section behaves @xmath37 as has been discussed at the of sect . 1
( see also ( [ eq : exp ] ) ) .
due to the baryon spin there are four gpds and hence , four form factors of which one decouples in the symmetric frame .
thus , the differential cross section for the production of a @xmath92 reads @xcite @xmath93 where @xmath94 the axial and the pseudoscalar form factors can be disentangled with polarization measurements . as for the meson case ( see ( [ eq : annff ] ) ) each form factor is a sum of individual flavor form factors ( @xmath95 ) @xmath96 where @xmath97 is a baryon - antibaryon distribution amplitude . for comparison the magnetic form factor of
the proton reads @xmath98 in fig .
[ fig : pp ] the belle data @xcite on @xmath99 are compared to the handbag results @xcite .
the effective and the vector form factors are parameterized analogously to ( [ eq : parameter ] ) and the parameters fitted to the data . from fig .
[ fig : pp ] one sees that the data are nicely compatible with a @xmath100 behavior .
the energy dependence of the form factors is @xmath101 , i.e. the integrated cross section falls off as @xmath102 at large @xmath64 .
note that in the erbl approach the cross section behaves @xmath103 . as for the case of mesons
one can show that due to the absence of @xmath104 states and @xmath55-spin symmetry there are only three independent flavor form factors for the ground state baryon channels for which one may take the ones for the proton - antiproton channel , @xmath105 .
the quality of the present data @xcite necessitates a simplifying assumption @xmath106 which leads to @xmath107 with @xmath108 a fit to the data @xcite provides the results shown in fig.[fig : pp ] ( @xmath109 , @xmath110 ) .
similar results are obtained for the @xmath111 channel . due to time - reversal invariance
the amplitudes for @xmath112 and @xmath113 are the same up to a phase .
this offers the opportunity for the fair project to study also handbag physics in the wide - angle region .
in addition to @xmath113 one may study the photon - meson channels .
their amplitudes are still under control of the basic form factors @xmath105 .
thus , for instance , for the @xmath114 channel the annihilation form factors read @xcite ( @xmath115 ) @xmath116 with these form factors one finds good agreement with the fermilab data @xcite on @xmath117 . finally , it should mentioned that the handbag approach to wide - angle scattering also applies to space - like processes .
the gold - plated example is real compton scattering ( cf . the remarks in sect .
[ intro ] ) for which the gpds or respective form factors are known from an analysis of the nucleon form factors @xcite .
the predictions for real compton scattering are in remarkable agreement with experiment .
the basic ideas of the handbag approach for wide - angle scattering are reviewed and applications to several time - like processes such as two - photon annihilations into pairs of mesons or baryon - antibaryons , are discussed .
it turns out that this approach seems to work quite well for energies larger than about 3 gev .
data at higher energies are highly welcome in order to have a clean suppression of subleading terms . in
the near future proton - antiproton annihilations into two photons or photon - meson can be measured at fair .
such data in combination with @xmath118 measurements performed at belle ii would allow for a detailed investigation of the proton - antiproton distribution amplitudes .
99 m. diehl , t. gousset , b. pire and o. teryaev , phys .
lett . *
* 8**1 , ( 1998 ) 1782
. m. diehl , p. kroll and c. vogt , phys .
b * * 5**32 , ( 2002 ) 99 .
m. diehl , p. kroll and c. vogt , eur .
j. c * * 2**6 , ( 2003 ) 567 .
a. freund , a. v. radyushkin , a. schafer and c. weiss , phys .
lett . * * 9**0 , ( 2003 ) 092001 .
m. diehl and p. kroll , phys .
b * * 6**83 , ( 2010 ) 165 .
s. j. brodsky and g. p.
lepage , phys .
d * * 2**4 , ( 1981 ) 1808 .
m. diehl , t. feldmann , r. jakob and p. kroll , eur .
j. c * 8 * , ( 1999 ) 409 .
h. nakazawa _ et al . _ [ belle collaboration ] , phys .
b * * 6**15 , ( 2005 ) 39 .
s. uehara _ et al . _ [ belle collaboration ] , phys .
d * * 7**9 , ( 2009 ) 052009 .
w. t. chen _ et al .
_ [ belle collaboration ] , phys .
b * * 6**51 , ( 2007 ) 15 .
s. uehara _ et al . _ [ belle collaboration ] , phys .
d * * 8**0 , ( 2009 ) 032001 .
s. uehara _ et al . _ [ belle collaboration ] , phys .
d * * 8**2 , ( 2010 ) 114031 .
v. l. chernyak , chin .
c * * 3**4 , ( 2010 ) 822 .
v. m. braun _ et al .
_ , phys .
d * * 7**4 , ( 2006 ) 074501 . c. -c .
kuo _ et al .
_ [ belle collaboration ] , phys .
b * * 6**21 , ( 2005 ) 41 .
p. kroll and a. schfer , eur .
j. a * * 2**6 , ( 2005 ) 89 . k. abe _ et al . _ [ belle collaboration ] , hep - ex/0609048 .
p. achard _ et al . _
[ l3 collaboration ] , phys .
b * * 5**36 , ( 2002 ) 24 .
t. a. armstrong _ et al . _
[ fermilab e760 collaboration ] , phys . rev .
d * * 5**6 , ( 1997 ) 2509 .
m. diehl , t. .feldmann , r. jakob and p. kroll , eur .
j. c * 39 * , 1 ( 2005 ) . | in this talk the handbag approach to hard exclusive wide - angle processes is reviewed and applications , as for instance two - photon annihilations into pairs of mesons , are discussed .
+ talk presented at meson 2012 , cracow , 2012 .
wu b 12 - 17 + september , 9 2012 + * hard exclusive wide - angle processes * + p. kroll + + and + |
the bilateral filter was proposed by tomasi and maduchi @xcite as a non - linear extension of the classical gaussian filter .
it is an instance of an edge - preserving filter that can smooth homogenous regions , while preserving sharp edges at the same time .
the bilateral filter has diverse applications in image processing , computer vision , computer graphics , and computational photography @xcite .
we refer the interested reader to @xcite for a comprehensive survey of the working of the filter and its various applications .
the bilateral filter uses a spatial kernel along with a range kernel to perform edge - preserving smoothing . before proceeding further
, we introduce the necessary notation and terminology .
let @xmath0 be a vector - valued image , where @xmath1 is some finite rectangular domain of @xmath2 .
for example , @xmath3 for a color image .
consider the kernels @xmath4 and @xmath5 given by @xmath6 where @xmath7 and @xmath8 is some positive definite covariance matrix .
the former bivariate gaussian is called the spatial kernel , and the latter multivariate gaussian is called the range kernel @xcite .
the output of the bilateral filter is the vector - valued image @xmath9 : \omega \rightarrow \mathbb{r}^d$ ] given by @xmath10(\i ) = \frac{1}{z(\i ) } \sum_{\j } \omega ( \j ) \ \phi\ !
\left ( \f(\i-\j ) - \f(\i ) \right ) \f(\i-\j),\ ] ] where @xmath11 the bilateral filter was originally proposed for processing grayscale and color images corresponding to @xmath12 and @xmath3 respectively @xcite . in practice , the sums in and are restricted to a square window @xmath13 ^ 2 $ ] around the pixel of interest , where @xmath14 is the standard deviation of the spatial gaussian @xcite .
thus , the direct computation of and requires @xmath15 operations per pixel .
in fact , the direct implementation is known to be slow for practical settings of @xmath14 @xcite . for the case
@xmath12 ( grayscale images ) , researchers have come up with several fast algorithms based on various forms of approximations @xcite .
a detailed account of some of the recent fast algorithms , and a comparison of their performances , can be found in @xcite .
a straightforward way of extending the above fast algorithms to vector - valued images is to apply the algorithm separately on each of the @xmath16 components .
the output in this case will generally be different from that obtained using the formulation in . in this regard , it was observed in @xcite that the component - wise filtering of rgb images can often lead to color distortions .
it was shown that such distortions can be avoided by applying in the cie - lab color space , where the covariance @xmath8 is chosen to be diagonal . in this paper
, we present a fast algorithm for computing .
the core idea is that of using raised - cosines to approximate the range kernel @xmath17 .
this approximation was originally proposed in @xcite for deriving a fast algorithm for gray - scale images .
it was later shown in @xcite that the raised - cosine approximation can be extended for performing high - dimensional filtering using the product of one - dimensional approximations .
unfortunately , this did not lead to a practical fast algorithm .
the fundamental difficulty in this regard is the so - called `` curse of dimensionality '' .
namely , while a raised - cosine of small order , say @xmath18 , suffices to approximate a one - dimensional gaussian , the product of such approximations result in an order of @xmath19 in dimensions @xmath20 .
a similar bottleneck arises in the context of computing using the raised - cosine approximation .
nevertheless , we will demonstrate how this problem can be circumvented using monte carlo approximation @xcite .
the contribution and organization of the paper are as follows . in section [ pa ] ,
we extend the shiftable approximation in @xcite for the bilateral filtering of vector - valued images given by . in this direction
, we propose a stochastic interpretation of the raised - cosine approximation , and show how it can be made practical using monte carlo sampling . based on this approximation
, we develop a fast algorithm in section [ fa ] . as an application
, we use the proposed algorithm for filtering color images in section [ results ] .
the results reported in this section demonstrate the accuracy of the approximation , and the speedup achieved over the direct implementation .
we conclude the paper in section [ conclusion ] . over the dynamic range
@xmath21 $ ] , where @xmath22 .
the raised cosine is of order @xmath23 , and @xmath24 for the monte carlo approximation ( obtained using a single realization of @xmath25).,scaledwidth=45.0% ]
as a first step , we diagonalize the quadratic form in the definition of @xmath17 using an orthogonal transform . indeed ,
since @xmath26 is positive definite , we can find an orthogonal matrix @xmath27 and positive numbers @xmath28 such that @xmath29 where @xmath30
. the change - of - basis @xmath31 amounts to transforming the components of the vector - valued image at each pixel ( similar to a color transformation ) .
in particular , by defining the image @xmath32 given by @xmath33 we can write as @xmath34(\i ) = \frac{1}{z(\i ) } \sum_{\j } \omega ( \j ) \
\psi \!\left ( \g(\i-\j ) - \g(\i ) \right ) \f(\i-\j),\ ] ] where @xmath35 comparing and the range kernel in , we see that @xmath36 is given by @xmath37 the key point here is that we have factored the original kernel @xmath17 into a product of one - dimensional gaussians . at this point ,
we recall the following result from @xcite .
[ prop1 ] @xmath38^n\ ! = \exp \!\left(- \frac{1}{2 } \alpha^2 t^2 \right).\ ] ] we apply proposition [ prop1 ] to the one - dimensional gaussians in , and conclude that @xmath39^n \ ! \ !
\!=\lim_{n \rightarrow \infty } \prod_{k=1}^d \left[\cos\!\left ( \frac{\alpha_k y_k}{\sqrt{n}}\right)\right]^n\!\!.\ ] ] in practice , we fix some @xmath40 , and replace with @xmath41^n.\ ] ] to reduce unnecessary symbols , we have used @xmath42 to represent both the original kernel and its approximation .
this should not lead to a confusion , since we will not use the original kernel in the rest of the discussion .
we will refer to @xmath40 as the _ order _ of the approximation .
the central observation of the paper is the following stochastic interpretation of , which is obtained by turning a product - of - sums into a sum - of - products .
[ prop2 ] let @xmath43 be a random vector , whose components are independent and follow the binomial distribution @xmath44 .
then @xmath45 where @xmath46 and the expectation @xmath47 $ ] is with respect to @xmath48 .
recall the identity @xmath49 , where @xmath50 .
we use this along with the binomial theorem , and get @xmath51 notice that the coefficient in corresponding to a given @xmath52 is simply the probability that a random variable @xmath53 takes on the value @xmath52 . in other words , we can write @xmath54 where @xmath47 $ ] denotes the mathematical expectation with respect to @xmath55 . substituting in , and using the independence assumption on the components of @xmath48 , we arrive at .
we next approximate the mathematical expectation in using monte carlo integration @xcite .
more specifically , let @xmath56 denote the distribution of the random vector @xmath57 in proposition [ prop2 ] that takes values in @xmath58 .
we fix a certain number of trials @xmath59 , and draw @xmath60 from @xmath58 using the distribution @xmath56 .
this gives us the following monte carlo approximation of : @xmath61 where @xmath62 is the @xmath63-th component of @xmath64 . to reduce symbols ,
we have again overloaded the notation for the original kernel in . in summary ,
starting from the kernel in , we arrive at the approximation in in two steps .
first , we approximate the one - dimensional gaussian using a raised - cosine of fixed order @xmath40 .
this results in a product of raised - cosines , which we express as a large sum - of - products . in the second step ,
we approximate the large sum using monte carlo integration .
a comparison of the raised - cosine and the monte carlo approximations for a one - dimensional ( @xmath12 ) range kernel is provided in figure [ kernelapproximation ] . before proceeding further
, we simplify the expression in by defining the random variable @xmath65 given by @xmath66 in terms of the realizations @xmath67 , we can write as @xmath68
we now present the fast algorithm obtained by using in place of the original gaussian kernel .
the fast algorithm is based on the so - called _ shiftability _ property of a function @xcite .
the shiftability of the exponential function follows from a rather simple fact , namely that @xmath69 . in particular , on substituting in , we can express the term @xmath70 as follows : @xmath71 where @xmath72 denotes the @xmath63-th component of @xmath73 .
we can simplify by introducing the ( complex - valued ) image @xmath74 given by @xmath75 and the image @xmath76 given by @xmath77 in term of , note that we can write as @xmath78 where @xmath79 denotes the complex - conjugate of @xmath80 . substituting in , and exchanging the two sums
, we obtain @xmath81 where the vector - valued image @xmath82 is given by @xmath83 .
similarly , on substituting in , we obtain @xmath84 notice that we have managed to express and using a series of gaussian convolutions . indeed , the term within brackets in is a gaussian convolution , which we denote by @xmath85 . on the other hand ,
the gaussian convolution in is performed on each of the @xmath16 components of @xmath86 ; we denote this using @xmath87 .
thus , we are required to perform a total of @xmath88 gaussian convolutions , which constitutes the bulk of the computation .
the procedure for computing the proposed approximation of is summarized in algorithm [ algo ] .
we shall henceforth refer to the proposed algorithm as the ` mcsf ` ( monte carlo shiftable filter ) .
note that we can efficiently compute the gaussian convolutions in step [ conv ] using separability and recursion , and importantly , at @xmath89 cost with respect to @xmath14 @xcite .
the run - time of the proposed algorithm is thus independent of @xmath14 , and is hence expected to be much faster than the direct implementation for large @xmath14 .
needless to mention , we can replace the spatial filter @xmath90 in with any arbitrary spatial filter ( such as a box or hat filter ) , and the above reductions would still hold . diagonalize @xmath8 to get @xmath91 and @xmath28 set up the spatial filter @xmath90 in using @xmath14 set @xmath92(\i)= z[\i]^{-1 } \mathbf{p}(\i)$ ] for @xmath93 .
in this section , we present some results on natural color images for which @xmath3 .
in particular , we demonstrate that the proposed ` mcsf ` algorithm is both fast and accurate for color images in relation to the direct implementation . to quantify the approximation accuracy for a given color image @xmath94
, we used the mean - squared error between @xmath9 $ ] and @xmath92 $ ] ( the latter is the output of algorithm [ algo ] ) given by @xmath95_k(\i)-\mathcal{s}[\f]_k(\i ) \big)^2,\ ] ] where @xmath9_k$ ] and @xmath92_k$ ] are the @xmath63-th color channel . on a logarithmic scale , this corresponds to @xmath96 db . for the experiments reported in this paper , we used an isotropic gaussian kernel corresponding to @xmath97 in .
in other words , @xmath98 , and @xmath99 for @xmath100 .
the accuracy of the proposed algorithm is controlled by the order of the raised - cosine ( @xmath40 ) and the number of trails ( @xmath59 ) .
it is clear that we can improve the approximation accuracy by increasing @xmath40 and @xmath59 .
we illustrate this point with an example in figure [ errorplot ] .
we notice that ` mcsf ` can achieve sub - pixel accuracy when @xmath101 and @xmath102 .
we have noticed in our simulations that , for a fixed @xmath59 , the accuracy tends to saturate beyond a certain @xmath40 .
this is demonstrated in figure [ errorplot ] using @xmath103 and @xmath23 .
.comparison of the run - time of the direct implementation and algorithm [ algo ] on the @xmath104 _ peppers _ image .
the range parameter used is @xmath105 .
the parameters of ` mcsf ` are @xmath101 and @xmath106 . the computations were performed using matlab @xmath107 on a @xmath108 ghz intel @xmath109-core machine with @xmath110 gb memory . [ cols="^,^,^,^,^,^,^",options="header " , ] [ table1 ] a comparison of the run - time of the direct implementation of and that of the proposed algorithm is provided in table [ table1 ] .
we notice that ` mcsf ` is few orders faster than the direct implementation , particularly for large @xmath14 .
indeed , following the fact that the convolutions in step [ conv ] of algorithm [ algo ] can be computed in constant - time with respect to @xmath14 @xcite , our algorithm has @xmath89 complexity with respect to @xmath14 . as against this , the direct implementation scales as @xmath15 .
finally , we present a visual comparison of the filtering for rgb images in figures [ dome ] and [ peppers ] .
notice that the outputs are visually indistinguishable .
as mentioned earlier , the authors in @xcite have observed that the application of the bilateral filter in the rgb color space can lead to color leakage , particularly at the sharp edges .
the suggested solution was to perform the filtering in the cie - lab space @xcite . in this regard ,
a comparison of the filtering in the cie - lab space is provided in figure [ house ] . in this case , we first performed a color transformation from the rgb to the cie - lab space , performed the filtering in the cie - lab space , and then transformed back to the rgb space .
the filtered outputs are seen to be close , both visually and in terms of the mse .
we proposed a fast algorithm for the bilateral filtering of vector - valued images .
we applied the algorithm for filtering color images in the rgb and the cie - lab space .
in particular , we demonstrated that a speedup of few orders can be achieved using the fast algorithm without introducing visible changes in the filter output ( the latter fact was also quantified using the mse ) .
an important theoretical question arising from the work is the dependence of the order and the number of trials on the filtering accuracy . in future work
, we will investigate this matter , and also look at various ways of improving the monte carlo integration @xcite .
we also plan to test the algorithm on other vector - valued images . | in this paper , we consider a natural extension of the edge - preserving bilateral filter for vector - valued images .
the direct computation of this non - linear filter is slow in practice .
we demonstrate how a fast algorithm can be obtained by first approximating the gaussian kernel of the bilateral filter using raised - cosines , and then using monte carlo sampling .
we present simulation results on color images to demonstrate the accuracy of the algorithm and the speedup over the direct implementation . bilateral filter , vector - valued image , color image , monte carlo method , approximation , fast algorithm . |
recently , social network has been a popular tool to share images .
when a social network user uploads an image , the image is usually associated with a tag/ keyword which is used to describe the semantic content of this image .
the tags provided by the users are usually incomplete .
zhang et al . designed and implemented a fast motion detection mechanism for multimedia data on mobile and embedded environment @xcite .
recently , the problem image tag completion is proposed in the computer vision and machine learning communities to learn the missing tags of images @xcite .
this problem is defined as the problem of complete the missing elements of a tag vector of a given image automatically . in this paper , we investigate the problem of image tag completion , and proposed a novel and effective algorithm for this problem based on local linear learning .
we propose a novel and effective tag completion method . instead of completing the missing tag association elements of each image
, we introduce a tag scoring vector to indicate the scores of assigning the image to the tags in a given tag set .
we propose to study the tag scoring vector learning problem in the neighborhood of each image .
for each image in the neighborhood , we propose to learn a linear function to predict a tag scoring vector from a visual feature vector of its corresponding image feature .
we propose to minimize the perdition error measure by the squared @xmath0 norm distance over each neighborhood , and also minimize the squared @xmath0 norm of the linear function parameters .
besides the local linear learning , we also proposed to regularize the learning of tag scoring vectors by the available tags of each image .
we construct a unified objective function to learn both the tag scoring vectors and the local linear functions .
we develop an iterative algorithm to optimize the proposed problem . in each iteration of this algorithm
, we update the tag scoring vectors and the local linear function parameters alternately .
this rest parts of paper are organized as follows : in section [ sec : method ] , we introduced the proposed method . in section
[ sec : exp ] , we evaluate the proposed methods on some benchmark data sets . in section [ sec : conclusion ] , the paper is concluded with future works .
we assume that we have a data set of @xmath1 images , and their visual feature vectors are @xmath2 , where @xmath3 is the @xmath4-dimensional feature vector of the @xmath5-th image .
we also assume that we have a set of @xmath6 unique tags , and a tag vector @xmath7^\top \in \{+1,-1\}^m$ ] for the @xmath5-th image @xmath8 , where @xmath9 if the @xmath10-th tag is assigned to the @xmath5-th image , and @xmath11 , otherwise . in real - world applications , the tag vector of an image @xmath8 is usually incomplete , i.e. , some elements of @xmath12 are missing .
we define a vector @xmath13\in \{1,0\}^m$ ] , where @xmath14 if @xmath15 is available , and @xmath16 if @xmath15 is missing .
we propose to learn a tag scoring vector @xmath17\in \mathbb{r}^m$ ] , where @xmath18 is the score of assigning the @xmath10-th tag to the @xmath5-th image .
the set of the @xmath19 nearest neighbor of each image @xmath8 is denoted as @xmath20 , and we assume that the tag scoring vector @xmath21 of a image @xmath22 can be predicted from its visual feature vector @xmath23 using a local linear function @xmath24 , @xmath25 where @xmath26 is the parameter of the local linear function . to learn the tag scoring vector and the local function parameters , we propose the following minimization problem , @xmath27 where @xmath28 and @xmath29 are tradeoff parameters . the objective function @xmath30 in ( [ equ : obj3 ] ) is a summarization of three terms over all the images in the data set .
the first term , @xmath31 , is the prediction error term of the local linear predictor over the neighborhood of each image .
the second , @xmath32 , is to reduce the complexity of the local linear predictor .
the last term , @xmath33 , is a regularization term to regularize the learning of tag scoring vectors by the incomplete tag vectors , so that the available tags are respected . to optimize the minimization problem in ( [ equ : obj3 ] )
, we propose to use the alternate optimization strategy @xcite in an iterative algorithm . * * optimization of @xmath34 * in each iteration , we optimize @xmath34 one by one , and the minimization of ( [ equ : obj3 ] ) with respect to @xmath35 can be achieved with the following gradient descent update rule , + @xmath36 + where @xmath37 is the sub - gradient function of @xmath38 , with respect to @xmath35 , + @xmath39 + and @xmath40 is the descent step . * * optimization of @xmath41 * in each iteration , we also optimized @xmath41 one by one . when @xmath42 is optimized , @xmath43 are fixed .
gradient descent method is also employed to update @xmath42 to minimize the objective in ( [ equ : obj3 ] ) , + @xmath44 + where @xmath45 is the sub - gradient function with respect to @xmath42 , + @xmath46
in the experiments , we used two publicly accessed image - tag data sets , which are corel5k data set @xcite and iapr tc12 data set @xcite . in the corel5k data set , there are 4,918 images , and 260 tages .
we extract density feature , harris shift feature , harris hue feature , rgb color feature , and hsv color feature as visual features for each image .
moreover , we remove 40% of the elements of the tag vectors to make the incomplete image tag vectors . in the iapr
tc12 data set , there are 19,062 images , and 291 tags .
we also remove 40% elements of the tag elements to construct the incomplete tag vectors .
to evaluate the tag completion performances , we used the recall - precision curve as performance measure .
we also use mean average precision ( map ) as a single performance measure .
we compared the proposed method to several state - of - the - art tag completion methods , including tag matrix completion ( tmc ) @xcite , linear sparse reconstructions ( lsr ) @xcite , tag completion by noisy matrix recovery ( tcmr)@xcite , and tag completion via nmf ( tc - nmf ) @xcite .
the experimental result on two data sets are given in fig .
[ fig : corel5k ] and fig .
[ fig : iapr ] . from these figures
, we can see that the proposed method loctc performs best .
its recall - precision curve is closer to the top - right corner than any other methods , and its map is also higher than maps of other methods . in this section , we will study the sensitivity of the proposed algorithm to the two parameters , @xmath28 and @xmath29 .
the curves of @xmath28 and @xmath29 on different data sets are given in fig .
[ fig : corel_alpah ] and fig .
[ fig : iapr_alpah ] . from these figures
, we can see that the performances are stable to different valuse of both @xmath28 and @xmath29 .
in this paper , we study the problem of tag completion , and proposed a novel algorithm for this problem . we proposed to learn the tags of images in the neighborhood of each image .
a local linear function is designed to predict the tag scoring vectors of images in each neighborhood , and the prediction function parameter is learned jointly with the tag scoring vectors .
the proposed method is compared to state - of - the - art tag completion algorithms , and the results show that the proposed algorithm outperforms the compared methods . in the future , we will study how to incorporate these connections into our model and learn more effective tags . in this paper , we used one single local function for each neighborhood , and in the future , we will use more than than regularization to regularized the learning of tags @xcite , such as usage of wavelet functions to construct the local function @xcite .
moreover , correntropy can also be considered as a alternative loss function to construct the local learning problem @xcite . in the future
, we also plan to extend the proposed algorithm for completion of tags of large scale image data set by using high performance computing technology @xcite , and completion of tags of gene / protein functions of bioinformatics problems @xcite .
feng , z. , feng , s. , jin , r. , jain , a. : image tag completion by noisy matrix recovery .
lecture notes in computer science ( including subseries lecture notes in artificial intelligence and lecture notes in bioinformatics ) * 8695 lncs*(part 7 ) , 424438 ( 2014 ) huang , y. , liu , q. , zhang , s. , metaxas , d. : image retrieval via probabilistic hypergraph ranking . in : proceedings of the ieee computer society conference on computer vision and pattern recognition , pp . 33763383 ( 2010 ) li , t. , zhou , x. , brandstatter , k. , zhao , d. , wang , k. , rajendran , a. , zhang , z. , raicu , i. : zht : a light - weight reliable persistent dynamic scalable zero - hop distributed hash table . in : parallel & distributed processing ( ipdps ) , 2013 ieee 27th international symposium on , pp . 775787 ( 2013 )
lin , z. , ding , g. , hu , m. , wang , j. , ye , x. : image tag completion via image - specific and tag - specific linear sparse reconstructions . in : proceedings of the ieee computer society conference on computer vision and pattern recognition , pp .
16181625 ( 2013 ) liu , l. , li , h. , xue , y. , liu , w. : reactive power compensation and optimization strategy for grid - interactive cascaded photovoltaic systems .
ieee transactions on power electronics * 30*(1 ) , 188202 ( 2015 ) wang , h. , wang , j. : an effective image representation method using kernel classification . in :
2014 ieee 26th international conference on tools with artificial intelligence ( ictai 2014 ) , pp .
853858 ( 2014 ) wang , j. , wang , h. , zhou , y. , mcdonald , n. : multiple kernel multivariate performance learning using cutting plane algorithm . in : systems ,
man and cybernetics ( smc ) , 2015 ieee international conference on .
ieee ( 2015 ) wang , j. , zhou , y. , duan , k. , wang , j.j.y . , bensmail , h. : supervised cross - modal factor analysis for multiple modal data classification . in : systems ,
man and cybernetics ( smc ) , 2015 ieee international conference on .
ieee ( 2015 ) wang , k. , kulkarni , a. , zhou , x. , lang , m. , raicu , i. : using simulation to explore distributed key - value stores for exascale system services . in : 2nd greater chicago area system research workshop ( gcasr ) ( 2013 )
wang , k. , zhou , x. , chen , h. , lang , m. , raicu , i. : next generation job management systems for extreme - scale ensemble computing . in : proceedings of the 23rd international symposium on high - performance parallel and distributed computing , pp .
111114 ( 2014 ) wang , k. , zhou , x. , li , t. , zhao , d. , lang , m. , raicu , i. : optimizing load balancing and data - locality with data - aware scheduling . in : big data ( big data ) ,
2014 ieee international conference on , pp .
119128 ( 2014 ) wang , k. , zhou , x. , qiao , k. , lang , m. , mcclelland , b. , raicu , i. : towards scalable distributed workload manager with monitoring - based weakly consistent resource stealing . in : proceedings of the 24rd international symposium on high - performance parallel and distributed computing , pp .
219222 ( 2015 ) zhang , s. , huang , j. , li , h. , metaxas , d. : automatic image annotation and retrieval using group sparsity .
ieee transactions on systems , man , and cybernetics , part b : cybernetics * 42*(3 ) , 838849 ( 2012 ) zhao , d. , zhang , z. , zhou , x. , li , t. , wang , k. , kimpe , d. , carns , p. , ross , r. , raicu , i. : fusionfs : toward supporting data - intensive scientific applications on extreme - scale high - performance computing systems . in : big data ( big data ) , 2014 ieee international conference on , pp .
6170 ( 2014 ) zhou , x. , chen , h. , wang , k. , lang , m. , raicu , i. : exploring distributed resource allocation techniques in the slurm job management system .
illinois institute of technology , department of computer science , technical report ( 2013 ) | the problem of tag completion is to learn the missing tags of an image . in this paper , we propose to learn a tag scoring vector for each image by local linear learning .
a local linear function is used in the neighborhood of each image to predict the tag scoring vectors of its neighboring images .
we construct a unified objective function for the learning of both tag scoring vectors and local linear function parameters . in the objective ,
we impose the learned tag scoring vectors to be consistent with the known associations to the tags of each image , and also minimize the prediction error of each local linear function , while reducing the complexity of each local function .
the objective function is optimized by an alternate optimization strategy and gradient descent methods in an iterative algorithm .
we compare the proposed algorithm against different state - of - the - art tag completion methods , and the results show its advantages .
tagging , tag completion , local learning , gradient descent |
this is a continuing study of the narrow - line region ( hereafter nlr ) of the seyfert galaxy ngc4151 using data obtained with the gemini near - infrared integral field spectrograph ( nifs ) .
the data comprise spectra of the inner@xmath12pc@xmath13 , at a spatial resolution of@xmath14pc at the galaxy , covering the wavelength range@xmath15@xmath16 m at a spectral resolving power r@xmath17 . in a previous paper ( * ? ? ?
* hereafter paperi ) we have used these data to map the nlr intensity distributions of @xmath18 emission lines , as well as their ratios .
the main results were the distinct flux distributions and physical properties observed for the ionised , molecular and coronal gas .
the ionised gas is co - spatial with the ionisation bi - cone observed in the optical ] emission at position angle ( hereafter pa)@xmath19 @xcite , and seems to trace the outflow along the bi - cone @xcite . in the inner region
, the nir ionised - gas emission extends beyond the borders of the cone , which does not have a sharp apex as noted also by @xcite .
the h@xmath20 molecular gas intensity distribution , on the other hand , avoids the region of the bi - cone and seems to originate in the galaxy disc , while the coronal gas emission is barely resolved . in another recent work @xcite
, we have studied the unresolved nuclear continuum showing that its origin is emission by hot dust within @xmath44pc from the nucleus as expected from the dusty torus postulated by the unified model @xcite .
the origin of this structure is probably a dusty wind that originates in the outer parts of the accretion disk .
this wind is probably clumpy , as suggested by recent models @xcite , and necessary in order to allow the escape of radiation along the equatorial plane of the bi - cone in order to ionize the gas and produce the observed flux maps of paperi . in the present paper
we use the nifs data to map the nlr kinematics of ngc4151 .
although many papers have been devoted to such a study , most of them are based on long - slit spectroscopy obtained in the optical with the hubble space telescope ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the present study was performed in the near - ir , a waveband which is less affected by dust , and using integral field spectroscopy , which allows a full two - dimensional ( hereafter 2d ) coverage of the nlr kinematics . with our data and analysis
we aim to examine previous claims of acceleration of the gas along the nlr , quantify the mass outflow rate and the corresponding feedback power , as well as investigate the origin of the nlr gas . as ngc4151 harbors the closest bright agn ,
its nlr is one of the best suited for this type of study .
our approach for the analysis of the nlr kinematics of ngc4151 is as follows .
first we map the centroid velocity and velocity dispersion of the ionised , molecular and coronal gas , obtained from fits to the emission - line profiles and compare our results with those of previous studies .
then we use an alternative approach to map the nlr kinematics , made possible by integral field spectrographs : a `` velocity tomography '' of the emitting gas , obtained by slicing the line profiles in velocity bins of@xmath21kms@xmath6 , producing channel maps which provide a clearer view of the velocity distribution in the different gas phases .
we have grouped the channel maps together in a sequence of velocity bins , generating movies which can be recovered at the authors website.thaisa/ifu_movies/ngc4151 ] we adopt in this paper the same distance to ngc4151 as in paperi:@xmath22mpc , corresponding to a scale at the galaxy of@xmath23pcarcsec@xmath6 @xcite .
the paper is organised as follows . in [ data ]
we provide a quick description of the data , in [ results ] we present the centroid velocity and velocity dispersion measurements , in [ tomography ] we present the emission - line `` tomography '' , in [ discussion ] we discuss our results , in [ feed ] we estimate the mass outflow rate along the nlr and compare with the mass accretion rate and in [ conclusion ] we present our conclusions .
the observations comprise three pointings obtained with the gemini north telescope with the nifs instrument @xcite operating with the altair adaptive optics module , each covering a square field of view of@xmath24arcsec@xmath13 .
one of the nifs fields was centred on the ngc4151 nucleus and the others were offset by@xmath25arcsec along position angle@xmath26 , resulting in a total field of view of@xmath27arcsec@xmath13 , with the longest extent oriented approximately along the radio jet @xcite .
the spectra cover the z , j , h , and k spectral bands at two - pixel resolving powers of @xmath28 , @xmath29 , @xmath30 , and @xmath30 , respectively .
the angular resolution , as determined from the fwhm of the psf , ranges from @xmath31arcsec in the h and k spectral bands to @xmath32arcsec in the z and j bands .
this corresponds to a spatial resolution at the galaxy of @xmath33 pc at h and k bands , and @xmath34pc in the z and j bands .
more details about the data and the data reduction can be found in paperi , as well as flux measurements of 61 emission lines from three characteristic locations along the nlr . in the present paper ,
we concentrate on the kinematics measured for the strongest emission lines of the ionised and molecular gas , namely : ] @xmath35 m , @xmath36 m , ] @xmath37 m , pa@xmath38@xmath39
m , ] @xmath40 m and h@xmath20@xmath41 m .
we have also measured the kinematics for the coronal lines of ] @xmath42 m and ] @xmath43 m .
we first present and discuss the centroid velocity and velocity dispersion maps and then present the velocity tomography .
centroid velocity and velocity dispersion ( @xmath44 ) maps were obtained by fitting gaussians to the emission - line profiles .
the velocities were obtained from the central wavelength of the gaussians ( after subtraction of the systemic velocity of @xmath45kms@xmath6 , and applying a heliocentric correction of @xmath46kms@xmath6 ) while the velocity dispersions were obtained from the full width at half maximum ( hereafter fwhm ) of the gaussians as @xmath47fwhm@xmath48 . in many locations along the nlr of ngc4151 , the strongest ionised gas emission lines
those of ] and show two kinematic components , with a mean velocity separation between them ranging from @xmath49 to @xmath50kms@xmath6 .
we were successful in fitting two gaussian components to both these lines , although the line has a broad component and a nuclear absorption component which introduces uncertainties in the fitting procedure for regions close to the nucleus .
in a few regions , we observe double components also in ] , pa@xmath38 , and br@xmath51 .
elsewhere , these lines are broader at locations where double components are observed in ] and . this suggests that the ] , pa@xmath38 , and br@xmath51 emission lines are also double . the lower signal - to - noise ratio ( hereafter s / n ) of these lines , the presence of a broad component for the emission - lines close to the nucleus , and the fact that the ] emission line appears to be intrinsically broader contribute to blend together two distinct kinematic components .
double components have been previously observed in the ] @xmath52 emission - line profiles by @xcite , and @xcite .
@xcite using higher resolution ( r@xmath53 ) spectra , showed that in some locations there are even three or four components .
our lower spectral resolution did not allow us to resolve more than two components .
typical fits to the ] emission - line profile are shown in fig .
[ f - profile ] , corresponding to locations ( a , b , c , d and e ) identified in the ] radial velocity maps which are shown in fig .
[ f - doublevel ] . because the s / n is not high enough to allow the fit of each component with the necessary number of free parameters , we had to constrain the width of the two gaussian components to the same value .
the implication of this assumption is that there will be some uncertainty in the gas kinematics derived in this way , mainly in the velocity dispersions , as discussed bellow .
although in the next section we derive the kinematics in terms of channel maps , with no assumptions , we decided to include the kinematics in terms of two components in order to be able to compare our results with those of previous studies , in which this was the method used to derive the gas kinematics .
in addition , the separation in two components helps in the identification of the origin of the gas emission . in the top panels of fig .
[ f - doublevel ] we show velocity maps in the emission lines of ] and of regions with one kinematic component , after subtraction of the systemic velocity . in the middle and bottom panels we show the velocity maps of the regions with two kinematic components : the middle panel shows the high velocity component , corresponding to negative ( blueshifted ) velocities to the sw and positive ( redshifted ) velocities to the ne , while the bottom panel shows the low velocity component . for the ] emission , the lowest velocities are observed along the galaxy minor axis ( pa@xmath54 ) in the upper panel of fig.[f - doublevel ] ( single component ) , and in neighboring locations where two components were fit to the data .
this low velocity component , essentially at zero velocity , probably originates in the galaxy disc , while the redshifted and blueshifted velocity components originate in the outflow along the bi - cone .
the highest blueshifts are observed in the bottom part of the strip labelled c " in the middle left panel of fig.[f - doublevel ] , while the highest redshifts are observed in the region labelled b " . in order to better illustrate the magnitude of the velocity variations , and to investigate further the origin of the different gas components , we show in fig .
[ f - pv ] the velocities as a function of distance from the nucleus along pa@xmath55 .
the black dots represent the single components and the red dots represent the double components , while the labels allow the comparison with fig.[f - doublevel ] .
[ f - pv ] suggests the presence of a @xmath4 zero velocity component observed between @xmath56 and @xmath57 from the nucleus along pa@xmath58 , whose origin is probably the galaxy plane , as pointed out above . then there is a sequence of blueshifts at @xmath59kms@xmath6 to the sw ( positive distances in fig.[f - pv ] ) which probably originates in the far side of the approaching cone ( see fig.[f - geom ] and discussion in sec.[geom ] ) .
a similar sequence of redshifts at somewhat lower velocities , up to @xmath60kms@xmath6 to the ne can be identified with the near side of the receding cone .
the highest blueshifts to the sw and highest redshifts to the ne can then be identified with the near side of the approaching cone and the far side of the receding cone , respectively .
these highest velocities thus originate in the walls of the bi - cone that are most inclined relative to the line - of - sight .
velocity dispersion ( @xmath44 ) maps of the ] and emission lines for the single and two - component fits are shown in fig .
[ f - doublesig ] .
we note again that the measured velocity dispersions are upper limits to the intrinisic gas velocity dispersion in regions where a single gaussian component has been fit , possibly to multiple unresolved components .
for example , in region e in fig .
[ f - doublesig ] , the velocity dispersion is @xmath2150kms@xmath6 . but this large velocity dispersion is probably due to the blending of at least two components , as it surrounds a region with two components ( region d ) .
similarly , the measured values are only indicative of the gas velocity dispersion in regions where two components have been fit , as their widths have been constrained to the same value .
this is the case of region e , where there are a component at zero velocity and a blueshifted component , which do not necessary have the same velocity dispersion .
nevertheless , these @xmath44 values should not be too far off because they are of the order of the @xmath44 value of the zero velocity single component along the minor axis of the galaxy .
the lowest @xmath44 values for the single components are @xmath0 100kms@xmath6 .
these are seen mostly along the minor axis of the galaxy .
the low velocity dispersion suggests that the emitting gas in these regions is not disturbed and probably lies close to the galaxy plane . the velocity dispersion is larger in other regions
this could be due to turbulence caused by the outflow or to multiple velocity components .
the @xmath44 values for the double components are in some regions ( e.g. region d " ) almost as low as the lowest values of the single components , suggesting that at least one of the components originates in the galaxy plane .
regions of higher @xmath44 values in the two components are those labelled
b " in the ne outflow and the bottom part of the strip c " , where the highest blueshifts are seen in the centroid velocity maps , supporting an origin in the front side of the sw cone , as discussed above . the centroid velocity distribution of h@xmath20 emission lines is dominated by velocities close to systemic .
this emission is concentrated along the minor axis of the galaxy , whose orientation approximately coincides with that of the galaxy bar .
there are some redshifts to the ne mainly along the galaxy major axis and some blueshifts to the sw along both the galaxy major axis and the cone axis ( see fig.[f - chmh2 ] ) .
the velocity dispersions are mostly constant at @xmath480kms@xmath6 supporting an origin for this gas in the galaxy plane .
the coronal line flux distributions are just barely resolved and their centroid velocities are also close to systemic , while their velocity dispersions are @xmath4100kms@xmath6 .
we have also mapped the gas kinematics by integrating the fluxes within wavelength slices extracted along the emission - line profiles corresponding to velocity bins of @xmath61kms@xmath6 .
these channel maps " display the distributions of the gas emission at different velocities , which occur at different depths along the line of sight . in this sense , they can be described as velocities tomography " of the nlr using emission - line scanning .
the channel maps are shown in figs .
[ f - chmsiii ] , [ f - chmfeii ] , [ f - chmbrg ] and [ f - chmh2 ] for the emission lines of ] , ] , br@xmath51 and h@xmath20 , respectively .
channel maps in the other ionised gas emission lines , such as and pa@xmath38 , are similar to those in ] and thus are not shown here .
the lowest flux shown in the channel maps is @xmath62 times the root mean square variation of the surrounding sky flux . the bi - cone axis at pa@xmath55 as determined by @xcite
is represented by a dashed line in the figures , while the major axis of the galaxy at pa@xmath63 is represented by a continuous line .
the contours of a radio merlin image ( kindly made available by c. mundell ) have been overplotted on the central velocity bins ( velocities close to systemic ) . in the channel maps of the ] emission line , shown in fig .
[ f - chmsiii ] , the most blueshifted gas at @xmath64kms@xmath6 lies very close to the nucleus . for blueshifts of @xmath65kms@xmath6
the emission extends further out , and for lower velocities broadens in the shape of a projected cone extending from the bi - cone axis towards the galaxy major axis . at the slice corresponding to @xmath66kms@xmath6 there seems to be a spatial association between the line emission and the radio structure .
in particular , both of them broaden towards the major axis . at velocities close to zero , and up to redshifts
@xmath67kms@xmath6 , the emission is extended in all directions around the nucleus , evidencing the absence of a sharp cone apex .
the largest elongation of this emission occurs along the horizontal axis , which is approximately the orientation of the radio axis .
the overplotted radio contours show that , although there is not a detailed correspondence between the ] and radio image , the brightest regions of the ] image coincide with the two inner knots of the radio image . for redshifts larger than @xmath68kms@xmath6 ,
the gas is observed mostly to the ne in the opposite direction to that of the blueshifted gas , showing also a similar flux distribution : for the lowest redshifts the emission is distributed in a fan structure extending from the major axis of the galaxy to the bi - cone axis , while for the intermediate and the highest redshifts the emission is more elongated and concentrated closer to the bi - cone axis to the ne .
channel maps for the ] emission line are shown in fig .
[ f - chmfeii ] . in the high - velocity blueshifted channels
, this emission seems to be more collimated along the bi - cone axis than for other emission lines .
some blueshifts from @xmath69kms@xmath6 to zero velocity are observed also to the ne , unlike the slices in the other emission lines .
the flux distributions for velocities close to zero and redshifted gas are similar to those in ] , although the highest redshifts are as high as the highest blueshifts and are observed up to similar distances from the nucleus . in all velocity channels , there seems to be more structure in ] than in ] , in the form of knots of emission in closer association with the radio knots .
this is particularly clear at the velocity bins from @xmath69 to @xmath70kms@xmath6 : at @xmath66kms@xmath6 the strongest emission traces the `` flaring '' of the radio structure to sw towards the galaxy major axis ; between velocities of @xmath71 and @xmath72kms@xmath6 , the strongest emission approximately traces the two central radio knots and , for higher redshifts , there is an enhancement of the ] emission at @xmath73arcsec ne , at the location of another radio knot .
fig.[f - chmbrg ] shows the channel maps in the br@xmath51 emission line .
these show similar morphologies to those seen in ] , but with more structure .
this can be attributed , at least in part , to the better spatial resolution achieved by the adaptive optics system in the k band relative to the z band .
the channel maps in the h@xmath20 emission line , shown in fig .
[ f - chmh2 ] , are completely different from those of the ionised gas , in support of a distinct origin for the molecular gas , as already pointed out in .
most of the emission occurs in the @xmath0 zero velocity channels .
a small excess of blueshifts to the sw ( from @xmath74kms@xmath6 ) and similar redshifts to the ne suggest rotation around the minor axis .
the flux distributions show a hollow core around the nucleus , and beyond this region the emission surrounds the minor axis of the galaxy in the shape of two arcs extending up to @xmath75arcsec to nw and se of the nucleus .
a comparison with the channel maps in the ionised gas at the same velocity bins ( around zero velocity ) shows that the molecular gas emission avoids the innermost region of ionised gas , and the hollow structure can thus be explained by the presence of ionised gas surrounding the nucleus at zero velocity .
in the centroid velocity maps for the ionised gas emission ( fig.[f - doublevel ] ) , the lowest velocities are observed at the nucleus , while the highest are observed away from the nucleus .
this is the case for ngc4151 and also for 6 other seyfert galaxies for which we have recently obtained kinematic data with the ifu of the gemini multi - objetct spectrograph ( gmos ) @xcite .
the channel maps , on the other hand , show a different behavior : the highest velocities are usually observed at the nucleus @xcite . in the case of ngc4151 , blueshifts of @xmath76kms@xmath6 in ] and ] are observed at the nucleus and in a narrow region along the bi - cone axis extending from the nucleus almost to the outer edge of the nlr ( figs .
[ f - chmsiii ] and [ f - chmfeii ] ) .
lower velocity blueshifted gas appears to fill the bi - cone and extend into the region of the radio jet near the systemic velocity .
a similar structure is seen in redshifted gas with the emission at moderate redshifts appearing to fill the receding bi - cone .
emission at the highest redshifts of @xmath77kms@xmath6 occurs at the nucleus and in a localised region @xmath0 0@xmath109 to the ne of the nucleus that is offset from the bi - cone axis .
this morphology differs from that of the highest velocity blueshifted gas and may simply reflect the clumpy nature of the nlr emission .
we attribute the difference between the kinematics probed by the centroid velocity and channel maps to the fact that the centroid velocity probes the brightest emission , while the channel maps also probe fainter emission ( in the wings of the line profiles ) . in the vicinity of the nucleus
the brightest component is the one originating in the galactic disc , while away from the nucleus the brightest component is the outflowing one . as a consequence ,
the centroid velocities show an apparent increase in velocity outwards , that mimics acceleration along the nlr . in the channel maps we see high velocity gas at the nucleus , showing that the outflow does not leave the nucleus at zero velocity .
we interpret this high velocity gas as an outflowing wind from the agn or ambient gas more directly interacting with this wind .
as this highest velocity gas moves away from the nucleus , it pushes and accelerates the gas from the disc .
the small amount of high velocity gas thus transfers momentum to a large reservoir of low velocity gas from the disc , causing the observed emission - line profiles and velocity structure along the bi - cone .
this scenario is consistent with the one proposed by @xcite who , after verifying the failure of various wind models to reproduce the observed outflows in ngc4151 , suggested that the interaction of an outflowing nuclear wind with the ism could explain the observed velocity profile .
some of the previous studies of the nlr of ngc4151 have suggested that there is no relation between the radio jet and the bi - conical outflow @xcite .
the distinct orientations of the two outflows do support distinct origins for them : the radio jet ( oriented at pa@xmath475@xmath3 ) originates in the funnel of the accretion disc , while the bi - conical outflow originates further out , probably still in the accretion disc .
the disc should also be warped , in order to produce a wind with a distinct orientation ( pa=60@xmath3 ) from that of the radio jet .
the channel maps of figs .
[ f - chmsiii ] , [ f - chmfeii ] and [ f - chmbrg ] confirm that the flux distributions at high velocities show no relation with the radio jet .
but at velocities close to systemic the flux distributions become oriented along the radio axis instead of along the bi - cone .
in addition , a comparison between the channel maps for velocities close to systemic and the contours of the radio map reveals line flux enhancements at the locations of radio knots .
our proposed interpretation is that the radio jet has been launched close to the plane of the galaxy , and does interact with the circumnuclear gas pushing it out approximately along the plane and introducing the flux enhancements mostly due to gas density enhancement which are observed at velocities close to systemic .
the observed flux enhancements could also be due to extra ionization by shock - induced x - rays in the nlr clouds interacting with the jet .
recent chandra x - ray observations of the nlr of ngc4151 @xcite have shown extended emission along the nlr with x - ray flux enhancements in association with radio knots .
they show that the radio - jet parameters are only consistent with the x - ray flux beeing due to thermal emission associated with the radio jet .
the presence of shock heating is supported by the results of paperi , namely the increase of the [ feii]1.257/[pii ] line ratio to values larger than @xmath42 in association with the radio knots .
values larger than @xmath42 imply the presence of fast shocks to destroy dust grains and release the fe .
we also find , in paperi , a high temperature ( 16 000k ) in the flux - enhanced regions , as derived from the [ feii ] line ratios , in agreement with this scenario . in the channel maps ,
the relation of the gas emission with the radio jet seems also to be strongest for ] , as can be observed in fig .
[ f - chmfeii ] : at velocities from @xmath78 to @xmath79kms@xmath6 , there is enhanced ] emission at the locations of most radio knots
. there seems also to be an association between a `` flared structure '' to the sw observed in the radio map with the line emission at velocities of up to @xmath80kms@xmath6 .
this suggests that the radio jet has been deflected towards the line - of - sight at this location .
some spatial association between optical and radio emission in the nlr of ngc4151 has also been observed by @xcite .
comparing high resolution radio images to an hst ] image , they found a number of bright ] clouds that seem to bound a number of knots observed in the radio image , which they interpret are due to shocks of the radio jet with the circumnuclear ism as the jet moves through the nlr and encounters gas clouds .
@xcite studied the ] kinematics of the nlr of the prototypical seyfert 2 galaxy , ngc1068 , using a long - slit spectrum obtained at pa 202 with stis on hst .
they found a distribution of line centroid velocities along the slit that resembled a `` figure of eight '' , with prominent high - velocity redshifted and blueshifted emission clumps and relatively few low - velocity clouds at intermediate distances from the nucleus .
this distribution led them to infer a highly inclined conical geometry for the nlr of ngc1068 , with the emission arising from clouds outflowing radially along the cone walls .
the cloud outflow velocities increase approximately linearly with radial distance on both sides of the nucleus to @xmath81kms@xmath6 at distances of @xmath82arcsec from the nucleus , and then decline back to zero by @xmath83arcsec .
@xcite were forced to hypothesise a gradual acceleration of the nlr clouds over @xmath84pc followed by gradual deceleration to @xmath85pc to account for this velocity structure .
@xcite used a more extensive set of stis spectra to develop this model for ngc1068 further .
@xcite applied the same model to stis spectra of the nlr of ngc4151 .
the nlr of ngc4151 is far less inclined to the line of sight than ngc1068
@xcite infer @xmath86 compared to @xmath87 for ngc1068 so radial flows along the cone walls do not separate as cleanly into high - velocity redshifted and blueshifted structures .
consequently , the fit to the ngc4151 data is less compelling than for ngc1068 .
@xcite revisited the ngc4151 nlr kinematics by fitting a similar model to spectra from five adjacent stis slit positions .
their comprehensive analysis provides strong evidence that the nlr of ngc4151 is inclined by @xmath88 to the line of sight , with a hollow bi - conical outflow occupying cone half - angles between @xmath89 and @xmath90 .
@xcite conclude that the outflow along the walls of the bi - cone is best reproduced with a velocity law : @xmath91 valid for distances @xmath92 from the nucleus up to a turnover point at @xmath93pc ( @xmath94 ) , from where the gas starts to decelerate and reaches the systemic velocity at @xmath95pc ( @xmath96 ) .
the sw side of the bi - cone is tilted toward us , as illustrated in fig .
[ f - geom ] . in this geometry and orientation
, we view the nucleus just outside the outer wall , at an angle of @xmath97 with respect to the nearest part of the sw cone .
this geometry and orientation has led to the idea that we see ngc4151 as a seyfert 1 galaxy because the blr can be viewed directly through a clumpy medium defining the cone walls .
the angle between the bi - cone axis and the normal to the galaxy disk is @xmath98 @xcite , and thus , for a total opening angle of the bi - cone of @xmath99 , one side of the cone exits the galactic disk steeply , with the outer wall making an angle of 87@xmath3 with the galaxy plane , while the other outer wall leaves the plane at only 21@xmath3 , impacting a larger area of the disk ( for a typical scale height for the disc of several tens of pc ) , as illustrated in fig.[f - geom ] .
@xcite argue that the intersection of the cone with the disc produces the observed geometry of the extended narrow - line region as shown by @xcite .
if we now inspect our centroid velocity maps of fig .
[ f - doublevel ] and fig.[f - pv ] in light of the geometry proposed by @xcite , we conclude that , in most regions where we see two velocity components , there is always one component at the systemic velocity plus an outflowing component in blueshift to the sw and in redshift to the ne .
we attribute the systemic velocity component to emitting gas from the galaxy disc , in which case we are led to the conclusion that the agn ionizes a region of the galaxy disc comparable in extent to that of the outflowing gas emission we detect along the bi - cone axis ( @xmath415 , or @xmath4100pc ) and somewhat less extended perpendicular to it ( @xmath41 , or @xmath465pc ) .
blueshifted components are observed to the sw at @xmath100kms@xmath6 and at higher velocities of up to @xmath5kms@xmath6 ( fig.[f - doublevel ] for ] and fig.[f - pv ] ) .
we interpret the component with blueshifts @xmath100kms@xmath6 at the position labelled _ d _ in fig.[f - doublevel ] , together with the single component surrounding it ( region _ e _ in fig.[f - doublevel ] ) , as originating in the back surface of the sw cone . in the bottom part of the region we labelled _ c _ in fig.[f - doublevel ] , we observe a few points with higher blueshifts of up to @xmath5kms@xmath6 ( see also fig.[f - pv ] ) .
these components probably originate in the front part of the sw cone , which we see at a small angle ( @xmath97 ) and thus with a large velocity component along the line - of - sight .
there is less emission from this front part of the sw cone than from the back , and our hypothesis to explain this is that the cone walls are formed by gas entrained from the galaxy disk .
the front wall is thus likely to be comprised of more tenuous gas than the back wall , as there is less gas to be entrained at high angles to the galactic disk .
redshifted components are observed to the ne at somewhat lower velocities than to the sw ( region labelled _
b _ in fig.[f - doublevel ] ) .
this can be seen more clearly in fig.[f - pv ] : there is a sequence of redshifts of up to @xmath60kms@xmath6 , which can be attributed to the front of the ne cone and then some higher redshifts of up to 400500kms@xmath6 , which can be attributed to the back of the ne cone . in fig.[f - doublevel ] , there is more emission from gas at the highest redshifts than emission from gas at the highest blueshifts .
this result , together with the lower redshifts to the ne than blueshifts to the sw , suggest that the ne cone has its axis somewhat tilted towards the line of sight , instead of being aligned with the axis of the sw cone .
this tilt would allow us to see more of the back wall of the ne cone and would result in lower observed velocities for both its front and back walls . in the model proposed by @xcite , the nlr
gas first accelerates up to @xmath4100pc from the nucleus and then decelerates , analogous to their successful model for ngc1068 .
while it is appealing in terms of agn unification for ngc1068 and ngc4151 to be fit with a similar model , neither our centroid velocity maps of figs.[f - doublevel ] and [ f - pv ] nor the ] kinematic data for ngc4151 unambiguously point to this interpretation .
the gradual acceleration over the inner @xmath101pc and then deceleration required by the model remain unexplained .
@xcite noted that radiative and magnetic acceleration from the region of the accretion disk both accelerate gas on length scales of the order of the launch radius , so winds driven by these mechanisms should reach their terminal velocities by @xmath102pc from the central source , not the @xmath103pc required by the model fits to the data .
thermal winds also fail in this regard @xcite .
dynamical models including various acceleration and deceleration mechanisms have been developed @xcite . while these models fail to identify the acceleration mechanism
, they do demonstrate that gravitational forces alone can not produce the required deceleration , while drag forces from the interaction of nlr clouds with a diffuse , hot , x - ray - emitting ambient medium could be responsible .
the difficulty in identifying a suitable wind acceleration mechanism leads to the question of whether some of the nlr kinematics are due to gradual global wind acceleration or whether they result from the local interaction of an already accelerated wind with its ambient medium , as we have proposed above on the basis of the channel maps .
this scenario has been proposed by @xcite , who realised that the interaction of an outflowing nuclear wind with the ism could reproduce the apparent acceleration of the nlr following the works of @xcite and @xcite .
these authors showed that a protostellar hydromagnetic wind expanding into the surrounding gas produces a bipolar outflow with velocity proportional to the distance to the star until a maximum radius beyond which the gas starts to decelerate . in order to be able to compare our data with those of previous works ( e.g. @xcite , @xcite ) ,
we have built position - velocity diagrams ( hereafter pv diagrams ) , which are shown in fig .
[ f - pv2siii ] for the ] emission line .
these diagrams were obtained from the @xmath104-band nifs data cube by extracting @xmath105arcsec wide pseudo - slit spectra oriented at pa@xmath55 ( which is the orientation of the bi - cone axis ) , offset from the nucleus perpendicular to the pseudo - slit direction by between @xmath106 ( along pa@xmath107 ) and @xmath108 ( along pa@xmath109 ) in steps of @xmath110 .
the ] diagrams are plotted both on a logarithmic scale ( top of fig .
[ f - pv2siii ] ) to reveal the faint emission structure and on a linear scale ( bottom ) to show the regions with strongest emission . while qualitatively similar to previous long - slit ] spectra ( e.g. fig . 2 of * ? ? ?
* ) , the ] diagrams seem to show less emission at velocities exceeding @xmath111kms@xmath6 . in particular , high - velocity ] clouds a and b in @xcite ( near @xmath112kms@xmath6 and @xmath113 ) are not apparent in our data . in part
, this could be due to the poorer spatial resolution obtained with nifs+altair at short near - infrared wavelengths . as with the ] line , most of the ] emission occurs at velocities below @xmath114kms@xmath6 .
the central diagrams of fig .
[ f - pv2siii ] can be compared directly with the ] slit spectrum shown in fig . 2 of @xcite .
the ] spectra are much less structured than the ] spectra , presumably due to poorer spatial resolution and consequent lower sensitivity to faint emission .
nevertheless , the two emission lines show a similar overall appearance along the central slit with blueshifted emission near @xmath115kms@xmath6 extending from @xmath116 to @xmath1173 from the nucleus towards pa@xmath118 ( sw , bottom of fig.[f - pv2siii ] ) , emission at the systemic velocity near the nucleus , and some redshifted emission near @xmath119kms@xmath6 extending to @xmath120arcsec from the nucleus towards pa@xmath55 ( ne , top of fig.[f - pv2siii ] ) .
the off - nucleus pv diagrams in fig .
[ f - pv2siii ] have a similar overall structure , but show that the @xmath115kms@xmath6 component is present in all negative offsets up to the slit at @xmath1218 .
of note is the fact that the velocity of this emission remains approximately constant over its full extent .
the brightest @xmath122kms@xmath6 emission is only clearly apparent towards pa@xmath55 along the cone axis ( i.e. , in the upper part of the central panel of fig .
[ f - pv2siii ] ) . some aditional emission from this component
can be observed along the slits with positive offsets at fainter levels and at larger distances from the nucleus .
the systemic velocity component is prominent in the upper section ( ne ) of pseudo - slits with negative offsets and in the lower sections ( sw ) of pseudo - slits having positive offsets , and originates in circumnuclear gas from the galaxy plane .
[ f - pvfeii ] shows the corresponding pv diagrams for the ] 1.644 line .
the emission in the central pseudo - slit corresponds closely to the brightest ] emission in fig . 2 of @xcite , but the high - velocity extensions of the ] clumps are much less prominent in our ] data .
although the major velocity components apparent in the ] line are also apparent in ] , the higher spatial resolution achieved at this longer wavelength helps to reveal sharper structures than in fig .
[ f - pv2siii ] . on the other hand ,
the structures seen in the ] emission can also in part be attributed to enhancements associated with the radio jet , as observed in the channel maps .
the component near @xmath115kms@xmath6 is comprised of several clumps and extends to only @xmath1235 along the central pseudo - slit towards pa@xmath118 , but reaches @xmath124arcsec at large negative , and possibly also positive , slit offsets i.e. , in the extreme left and right panels of fig .
[ f - pvfeii ] . clouds with velocities
@xmath125kms@xmath6 are observed at @xmath126arcsec along the central slit and closer to the nucleus along the @xmath110 and @xmath127 offset slits .
emission with velocities up to @xmath128kms@xmath6 is apparent at slit offsets of @xmath129 and @xmath130 .
the systemic - velocity component is restricted to the nucleus in the central slit , but extends to nearly @xmath131arcsec along the @xmath130 to @xmath132 offset slits and to nearly @xmath133arcsec along the @xmath129 and @xmath134 offset slits
. however , the systemic - velocity emission at negative slit offsets to the left in fig .
[ f - pvfeii ] is dominated by a further bright clump near @xmath133arcsec approximately east of nucleus , in the vicinity of the receding radio lobe .
overall the impression from the ] data is of roughly constant redshifted velocities towards pa@xmath55 and blueshifted velocities towards pa@xmath118 with an additional systemic - velocity component originating in the galaxy disc and associated with the radio jet .
the role of the geometry and projection effects on the observations can best be evaluated via modelling . in this section
, we compare our data with models with and without acceleration along the nlr .
we begin by testing the model proposed by @xcite on the basis of the optical ] emission obtained with stis , namely of an inclined hollow cone with radial outflow along the cone walls that is characterised by linear acceleration followed by linear deceleration @xcite .
our 3d kinematic data allow us to test this model against both channel maps and pv diagrams through multiple pseudo - slits in ways that were not available to previous authors .
we begin by presenting both these views for the model parameters of @xcite in figs .
[ f - daspv ] and [ f - daschm ] . as our data cover essentially the entire region of acceleration in the models , emission from decelerating gas beyond the velocity maximum has been suppressed in fig [ f - daschm ] . .
] the @xcite model predictions differ in several ways from the near - infrared emission - line data .
first , the high velocity gas just to the sw and ne of the nucleus predicted by the models ( fig .
[ f - daspv ] ) and seen in ] is less conspicuous in our data .
it is seen in our data only in the log plots for ] , where the faint emission gets enhanced ( top panel of fig.[f - pv2siii ] ) .
we also see little evidence of acceleration and deceleration in the near - infrared data , which mostly show sequences at constant velocities from the nucleus all the way out . even in the deep ] pv diagrams , gas up to @xmath135arcsec from the nucleus in the lower half of these diagrams extending to pa@xmath118 is still moving at blueshifted velocities of @xmath100kms@xmath6 , with similar but smaller redshifted velocities in the upper half of the diagrams .
another significant difference between the model and the near - infrared pv data is that the model predicts that pseudo - slits offset from the nucleus will show no emission near the slit centre fig .
[ f - daspv ] .
this is the equatorial region away from the nucleus and between the two oppositely - directed cone `` lobes '' .
most of the actual emission at off - nucleus slit positions arises in this region that is excluded by the model
c.f . , fig .
[ f - pvfeii ] .
there is also emission at zero velocity mainly observed in the ] pv diagram which is not included in the model . finally , the model predicts that the channel maps should show ridges of emission perpendicular to the cone axis at high positive and negative velocities fig .
[ f - daschm ] .
this morphology arises because outflowing material in the hollow cone essentially emits from an annular region perpendicular to the cone axis
convolved with other orientation factors .
in contrast , the channel maps in figs .
[ f - chmsiii ] and [ f - chmfeii ] show that the highest - velocity emission is confined to positions close to the cone axis , and is possibly extended along the cone axis towards the nucleus .
we now investigate modifications to the hollow - cone , radial - outflow model that attempt to address the above differences .
the simplest modification is to hypothesise that the cone wall that is steeply inclined to the plane of the sky does not emit strongly in near - infrared lines .
this could occur because this cone wall is more inclined to the galactic disk . as discussed in sec .
[ centroid ] , under the hypothesis that the cone walls are formed by gas entrained from the galaxy disk , this wall is likely to be comprised of more tenuous gas , as there is less gas to be entrained at high angles to the galactic disk .
this modification removes the high - velocity components in fig.[f - daspv ] . to this
, we then add an illustrative systemic - velocity component that is aligned with the radio jet and an arbitrary galactic disk component .
the radio - jet component is modeled as a central nuclear component plus two clumps offset e - w from the nucleus by 1 .
each clump is assumed to be at the systemic velocity of the galaxy .
the galaxy component has a gaussian intensity profile and is assumed to be at the systemic velocity .
the presence of this component is suggested both by the zero velocity component in the centroid velocity maps fig .
[ f - doublevel ] and by the channel maps , which show strong emission at zero velocities ( figs .
[ f - chmsiii ] and [ f - chmfeii ] ) .
the resulting pv diagrams are shown in fig.[f - modpv ] .
the pv diagrams in fig.[f - modpv ] do not provide a perfect fit to the data ( in particular to the highest spatial resolution ] data in fig.[f - pvfeii ] ) .
however , they serve to illustrate the following points : firstly , the emission from the cone ( blue regions in fig.[f - modpv ] ) is a better fit to the data in the linear plots ( showing the strongest emitting gas ) if only the section of the cone wall closest to the plane of the sky , and hence closest to the galactic disk , is included . this may be due to interaction of the nlr gas in the cone wall with the galactic disk , possibly by the entrainment of disk material into the nlr outflow . given that this material is moving in a direction closer to the plane of the sky , we are less sensitive to its true velocity so can not make strong claims about the length scale on which it is accelerated .
fig.[f - modpv ] is drawn for the gradual acceleration and subsequent deceleration proposed by @xcite .
however , the data permit virtually any velocity law , including constant velocity outflow , on the length scales to which they are sensitive . the second point illustrated by fig.[f - modpv ]
is that emission from gas interaction with the radio jet is significant .
this is modeled as nuclear and off - nuclear clumps , but in reality other emitting gas forms a bridge between the main clumps and the nucleus .
this is seen in the channel maps shown in fig.[f - chmfeii ] .
finally , the galaxy component at systemic velocity reproduces the data at pa=60@xmath3 for pseudo - slits at negative offsets and at pa@xmath136for pseudo - slits at positive offsets , supporting the presence of a disc component , as concluded also from the centroid velocity maps of fig.[f - doublevel ] .
simulated channel maps are shown in fig.[f - modchm ] .
these only approximately reproduce the data presented in fig.[f - chmfeii ] . in particular , two separated emission clumps aligned perpendicular to the cone axis are predicted from the cone walls at high - redshifted and high - blueshifted velocities while only a single clump near the cone axis is actually seen in the data .
.[f - modchm ] ] finally , we have considered a truncated hollow conical outflow at a constant velocity of @xmath137kms@xmath6 , as we could not see acceleration or deceleration in our data . as in the previous model , we add the radio jet and galactic disk component .
the resulting position - velocity diagrams are shown in fig .
[ f - modconstpv ] .
the fit to the observed constant velocity sequence at @xmath100kms@xmath6 for ] ( fig.[f - pv2siii ] ) along the slit passing through the nucleus and the ones with negative offsets are improved by this model .
the disc component at zero velocity reproduces emission observed at the nucleus and along pa@xmath138 ( positive position in the panels of fig .
[ f - pv2siii ] ) and pa@xmath139 ( to the left in fig .
[ f - pv2siii ] ) , as well as less extended emission along pa@xmath140 ( negative positions in fig .
[ f - pv2siii ] ) and pa@xmath141 ( to the right in fig .
[ f - pv2siii ] ) .
the high velocity sequences predicted by the models although not conspicuous in the linear pv diagrams of ] ( bottom panel of fig.[f - pv2siii ] ) are observable in the log pv diagrams ( upper panel of fig.[f - pv2siii ] ) , due to its faint emission .
these high velocity sequences and are due to the front wall of the approaching cone and far wall of the receeding cone .
we point out that this short high velocity sequence as predicted by the constant velocity model seems to be adequate to reproduce also the high velocity gas observed in ] ( fig . 2 of @xcite ) .
nevertheless , this model did not improve the fit for pseudo - slits offset by more than @xmath127 from the nucleus , which still do not show emission near the slit centre in the model .
in addition , these pseudo - slits show velocity sequences that decrease with increasing distance from the nucleus . in order to improve the fit ,
new geometries should be considered .
for example , a flow that is not conical , but in the form of a bowl as observed in planetary nebulae and in recent nifs observations of ngc1068 @xcite .
simulated channel maps are shown in fig .
[ f - modconstchm ] .
these maps are an improvement relative to the previous model for velocity channels between @xmath142kms@xmath6and @xmath143kms@xmath6 , but show double structures for velocities between @xmath144kms@xmath6and @xmath145kms@xmath6 , which are not seen in the data . for the highest velocities ,
a small triangular structure is seen , while the data show only clumps near the cone axis .
this discrepancy can also be attributed to possible deviations from a bi - conical geometry and to very weak emission from the wall of the bi - cone that is most inclined .
( see text ) .
the orientation and velocity bins are as for fig.[f - pv2siii ] .
[ f - modconstchm ] ] the modelled integrated emission - line map is similar for the three models discussed above .
we show in fig .
[ f : mod_int ] the map obtained for the modified @xcite model . a comparison with fig .
[ f : feii_int ] shows that it provides a reasonable reproduction of the data , taking into account that one should add the contribution of gas from the galaxy plane to the modelled emission .
@xcite identified blueshifted ] emission with velocities up to @xmath146kms@xmath6 in the ngc1068 nlr @xmath147pc ne of the nucleus .
this is the region attributed to the highest radial outflow velocities in the gradual acceleration model .
many of these knots are linked kinematically to more massive and slower - moving clumps located closer to the nucleus . @xcite
identify these knots as ablation flows from disintegrating molecular clouds that are being photoionised and radiatively accelerated by the agn .
@xcite detected faint , high - velocity ] emission with radial velocities up to @xmath148kms@xmath6 in stis slitless spectra of ngc4151 .
the high - velocity emission was shown to arise in high - excitation ionised gas , and has velocity dispersions in excess of @xmath149kms@xmath6 @xcite , which is also consistent with an origin in an ablation flow .
near - infrared ] emission arises in conditions similar to the optical ] emission , and can thus be formed in tenuous fully ionised gas experiencing a high local ionisation parameter .
but then , why do we see less ] than ] emission ?
either we lack the sensitivity to detect faint high - velocity emission , or this emission is not present . in order to investigate this , version 8.0 of the photoionisation code cloudy , last described by @xcite ,
was used to calculate the ratio of ] to ] for nlr clouds with a range of densities .
the ionisation parameter was set to @xmath150 for a gas density of @xmath151@xmath152 and then varied inversely with the gas density to simulate clouds with different densities being illuminated by the same radiation field .
such a situation would occur if the low - velocity ] is emitted at the front face of a dense nlr cloud and the high - velocity ] emission arises in an associated low - density ablation flow .
the resulting emission - line ratios are plotted in fig .
[ f - ratio ] where it can be seen that ] / ] increases from @xmath153 at densities typical of nlr clouds to @xmath154 at densities @xmath155 @xmath152 that might be typical of an ablation flow .
the values at high densities are typical of the integrated line ratios observed in seyfert galaxies nlrs ( e.g. , * ? ? ?
the strength of ] falls sharply to lower densities while the strength of ] is largely unchanged .
the reason for this dependence on ionisation parameter is that sulphur exists predominantly as ] at typical nlr densities , but as ] and higher species in more highly excited low - density regions .
in contrast , the higher ionisation potential of ] ( 54.9 ev compared to 35.1 ev for ] and 47.2 ev for ] ) means that ] survives better in highly excited regions .
the lower intensity of ] in low - density irradiated regions makes it more difficult to detect a low - density ablation flow in ] emission than in ] emission .
this combined with the lower strehl ratio and consequent poorer image quality achieved by adaptive optics systems at short near - infrared wavelengths probably account for our inability to detect more high - velocity gas in ngc4151 . there is little evidence in our near - infrared data for the gradual acceleration followed by gradual deceleration proposed by @xcite based on ] kinematics .
no viable physical mechanism has been identified for such a gradual acceleration .
a possible explanation is that the high - velocity ] emission does not represent the bulk motion of the outflow .
rather , this could be low - density material ablated from nlr clouds that we observe outflowing at @xmath4250kms@xmath6 and are accelerated much closer to the agn .
this is consistent with the low surface brightness of the high - velocity , low - density emission in ] and with the prominence of near - infrared emission formed on the sides of the outflow closer to high - density gas in the galactic disk .
while the kinematics of the ] , , , and ] emitting gas is dominated by outflows along the bi - cone , the h@xmath20 radial velocity field in fig .
[ f - chmh2 ] is dominated by velocities close to zero around the galaxy minor axis .
although the velocity range is small , it can be observed that , from the blueshifted to the redshifted channels , the flux distribution moves around the minor axis of the galaxy , from the sw to the ne , as expected for gas rotating in the galaxy plane .
although we see a kinematic component at systemic velocity also in the ionised gas , we do not see the outflowing component in h@xmath20 .
the velocity dispersion in the h@xmath20 gas is also the smallest among the emission - line profiles .
distinct kinematics for the h@xmath20 emitting gas relative to the emission lines of ionised gas has also been observed in previous studies of other objects @xcite .
the h@xmath20 emission lines usually have small velocity dispersions and their kinematics seem to be dominated by rotation in the galaxy plane , while the other emission lines have larger velocity dispersions and other velocity components besides rotation , as observed also in ngc4151 .
in paperi we have argued that the flux distributions , physical properties and location suggest that the h@xmath20 emission gas traces the agn feeding , while the ionised gas traces its feedback .
although we do not see the h@xmath20 emitting gas actually flowing down into the nucleus , it does present a distinct kinematic signature relative to the ionised gas .
the inflow may be occurring further in , in non - emitting colder molecular gas .
the hot molecular gas is estimated to be a fraction of only @xmath156 to @xmath157 of the total mass of molecular gas @xcite in agn . in the case of ngc4151 , for which we obtained a mass of hot h@xmath20 of @xmath158m@xmath8
, we can thus estimate a total molecular gas mass of@xmath159m@xmath8 .
such a large molecular gas reservoir around the nucleus of active galaxies is supported by radio observations of the central region of active galaxies in general , by , for example , the `` nuga '' group , which report molecular gas masses in the range @xmath160@xmath161 in the inner few hundred parsecs of active galaxies ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
if such a reservoir of molecular gas has accumulated by a flow that was the trigger of the present nuclear activity , it can be concluded that a steady flow would have to bring gas inwards at a rate in the range @xmath162@xmath163m@xmath164yr@xmath6 . in ngc4151 ,
@xcite have measured inflows in radio observations of that occur along narrow channels interpreted as shocks along the leading edges of the large scale bar ( at pa@xmath165 ) .
these inflows seem to end at the inner part of the bar in the nuclear region .
the orientation of the flow is approximately perpendicular to the `` arcs '' seen in the h@xmath20 flux distribution see fig .
[ f - chmh2 ] and one possibility is that the radio observations are tracing the gas inflows that accumulate in the nuclear region and contribute to build up the molecular gas , part of which we observe due to excitation of the h@xmath20 molecule by x - rays emitted by the nucleus .
the large amounts of molecular gas observed around the nuclei of active galaxies can only lead to the conclusion that inflows must be occurring @xcite .
nevertheless , the actual kinematic observation of inflowing gas in active galaxies is difficult to obtain .
this is due to the fact that the emission of the cold , inflowing gas is much weaker than the emission of the hotter , outflowing gas . in @xcite and
@xcite we were successful in mapping the inflow in low - excitation ionised gas because the targets were low - activity agn the liners ngc1097 and ngc6951 for which the nuclear radiation and outflows are much weaker than in seyfert galaxies , thus allowing the observation of inflows in the disc in ionised gas .
we are probably seeing just the ionised `` skin '' of a flow of cold gas on its way to feed the agn .
we were also successful in mapping an inflow observed in hot molecular gas in the galaxy ngc4051 @xcite . in this latter case ,
the x - ray emission of the agn may have worked as a flashlight which `` illuminated '' the inflowing molecular gas .
we obtained a mass inflow rate of @xmath166m@xmath164yr@xmath6 .
if we use the same ratio of hot to total molecular gas as above , the inflow rate becomes of the order of a few m@xmath164yr@xmath6 . another recent case of inflow seen in molecular gas was reported by @xcite , using near - ir ifu observations of ngc1068 obtained with sinfoni .
the feedback from the agn in the form of the mass outflow rate along the nlr and its kinetic power can be calculated using the ionised gas luminosity and geometry of the outflow through the expression : @xmath167 where @xmath168 is the mass of the proton , @xmath169 is the ionised gas density , @xmath170 is the outflow velocity , @xmath171 is the gas filling factor and @xmath172 is the cross section area of the outflow .
the filling factor was estimated from : @xmath173 where @xmath174 is the br@xmath51 luminosity emitted by a volume @xmath175 and @xmath176erg@xmath152s@xmath6 for a temperature t=10000k and gas density @xmath177@xmath152 @xcite . as in but got as a result f@xmath1781 , which is unphysical , supporting a higher average gas density . ]
the volume @xmath175 was calculated assuming the geometry proposed by @xcite and used in the modelling described in the previous section , namely a hollow bi - cone with a height of 130pc for each cone ( corresponding to 2 ) , and inner and outer opening angles of @xmath179 and @xmath180 , respectively .
we obtained @xmath181@xmath152 .
the br@xmath51 luminosity was obtained from paperi .
we have further divided this by a factor of 2 , because part of the ionised gas emission comes from galactic disc gas that is not outflowing ( see previous section ) .
we estimate from the channel maps that approximately half of the total ionised gas luminosity arises in this galactic disc gas . substituting these estimates in the expression above
we obtain a filling factor value of @xmath182 . filling factor values
reported in the literature are usually smaller , but they usually consider an isotropic ( spherical ) outflow . assuming an isotropic outflow instead of a bi - conical one , we obtain @xmath183 which is similar to the typical values in the literature .
the cross section area @xmath172 of the hollow cone was calculated at a distance of 65pc ( 1 ) from the cone apex for the hollow cone geometry and resulted in @xmath184cm@xmath13 . adopting the velocity @xmath186kms@xmath6 for the outflow ( which is the one assumed for the constant velocity outflow model ) , @xmath187@xmath152 , @xmath182 we finally obtain a mass outflow rate of @xmath188m@xmath164yr@xmath6 in each cone .
this value is of the order of the mass outflow rates reported for the nlr of active galaxies in neutral and ionised gas @xcite , and @xmath06 times larger than the mass outflow rate estimated by @xcite from uv absorbers in the spectra .
the mass outflow rate can be compared with the accretion rate to feed the supermassive black hole .
the accretion rate can be estimated from : @xmath189 where @xmath190 is the speed of light , and the bolometric luminosity is @xmath191ergs@xmath6 @xcite , which is 0.012 times the eddington luminosity for a black hole mass of @xmath192m@xmath8 @xcite . for this eddington ratio
we adopt an efficiency factor for the conversion of rest mass energy into radiation of @xmath193 , a typical value for an optically thin geometrically thick accretion disc @xcite .
the result is @xmath194yr@xmath6 , which is two orders of magnitude smaller than the mass outflow rate , indicating that the origin of the outflowing gas is not the agn but could be surrouding gas from the galaxy ism accelerated by a nuclear outflow , as discussed in the previous section .
our data also allows us to estimate the kinetic power of the outflow , including both the radial and turbulent components of the gas motion . assuming that the velocity dispersion of the hii emitting gas is due to turbulence , the kinetic power can be estimated as : @xmath195 from fig .
[ f - doublesig ] we obtain @xmath196kms@xmath6 , and adopting as above , @xmath197kms@xmath6and using @xmath198 ( considering the outflows in both cones ) , we obtain @xmath9ergs@xmath6 , which is @xmath199l@xmath200 . thus , from the power that can be extracted from the accretion rate of @xmath194yr@xmath6 , @xmath40.3% is transformed in kinetic power in the nlr outflows .
we note that this fraction is of the order of the agn feedback derived in simulations for co - evolution of black holes and galaxies ( 0.5% ) , in order to match the m-@xmath44 relation @xcite .
this kinetic power is somewhat larger than those we have recently derived for a sample of 6 nearby seyfert galaxies showing nlr outflows @xcite , which are in the range @xmath201 . in order to gauge the impact that the above kinetic power may have in the galaxy
, we estimate the total kinetic energy released if the outflow as now observed lasts for at least @xmath160yrs ( as an estimated duration for the activity phase of the galaxy ) to be @xmath202erg .
this energy can then be equated to the binding energy of the gas mass that could be pushed away from the central region of the galaxy ( following the reasoning in @xcite ) @xmath203 . for @xmath204erg , @xmath205m@xmath8 ( estimated from the galaxy absolute magnitude ) and @xmath206pc ( just beyond the observed outflow ) , we obtain @xmath207m@xmath8 , which is @xmath4 5 times the mass which outflows through the nlr in 10@xmath208yrs .
it is interesting to note that this mass is of the order of the estimated molecular gas mass present in the nuclear region of ngc4151 , as discussed above , being also the typical value observed in the nuclear region of active galaxies in general @xcite .
thus , it appears that in the case of ngc4151 the molecular gas reservoir that is needed to trigger the nuclear activity can be blown away by the nlr outflow by the end of the activity cycle .
we have measured the kinematics of the ionised and molecular gas surrounding the nucleus of ngc4151 using integral - field spectroscopy obtained with the nifs instrument at the gemini north telescope .
the main results of this paper are :
antonucci , r. r. & miller , j. s. , 1985 , apj , 297 , 621 | we have used the gemini near - infrared integral field spectrograph ( nifs ) to map the gas kinematics of the inner @xmath0 200@xmath1500pc of the seyfert galaxy ngc4151 in the z , j , h and k bands at a resolving power @xmath25000 and spatial resolution of @xmath08pc .
the ionised gas emission is most extended along the known ionisation bi - cone at position angle pa=60240@xmath3 , but is observed also along its equatorial plane .
this indicates that the agn ionizes gas beyond the borders of the bi - cone , within a sphere with @xmath41 radius around the nucleus .
the ionised gas has three kinematic components : ( 1 ) one observed at the systemic velocity and interpreted as originating in the galaxy disk ; ( 2 ) one outflowing along the bi - cone , with line - of - sight velocities between @xmath5 and 600kms@xmath6 and strongest emission at @xmath7kms@xmath6 ; ( 3 ) and another component due to the interaction of the radio jet with ambient gas . the radio jet ( at pa=75255@xmath3 ) is not aligned with the nlr , and produces flux enhancements mostly observed at the systemic velocity , suggesting that the jet is launched close to the plane of the galaxy ( @xmath0 plane of the sky ) . the mass outflow rate , estimated to be @xmath41m@xmath8yr@xmath6 along each cone
, exceeds the inferred black hole accretion rate by a factor of @xmath0100 .
this can be understood if the narrow - line - region ( nlr ) is formed mostly by entrained gas from the circumnuclear interstellar medium by an outflow probably originating in the accretion disk .
this flow represents feedback from the agn , estimated to release a kinetic power of @xmath9ergs@xmath6 , which is only @xmath00.3% of the bolometric luminosity of the agn .
there is no evidence in our data for the gradual acceleration followed by gradual deceleration proposed by previous modelling of the [ oiii ] emitting gas .
our data allow the possibility that the nlr clouds are accelerated close to the nucleus ( within 0@xmath101 @xmath46pc ) after which the flow moves at essentially constant velocity ( @xmath11kms@xmath6 ) , being consistent with nir emission arising predominantly from the interaction of the outflow with gas in the galactic disk .
the molecular gas exhiibits distinct kinematics relative to the ionised gas .
its emission arises in extended regions approximately perpendicular to the axis of the bi - cone and along the axis of the galaxy s stellar bar , avoiding the innermost ionised regions .
it does not show an outflowing component , being observed only at velocities very close to systemic , and is thus consistent with an origin in the galaxy plane .
this hot molecular gas may only be the tracer of a larger reservoir of colder gas which represents the agn feeding .
[ firstpage ] galaxies : nuclei , galaxies : active , galaxies : seyfert , galaxies : kinematics , galaxies : jets , galaxies : individual ( ngc4151 ) |
recently , the zeus and h1 experiments at hera have reported the first evidence for diffractive deeply inelastic electron scattering @xcite , @xmath0 this is an example of a more general phenomenon , diffractive hard scattering , in which a high energy incident hadron participates in a hard interaction , involving very large momentum transfers , but nevertheless the hadron itself is diffractively scattered , emerging with a small transverse momentum and the loss of a rather small fraction of its longitudinal momentum .
one may say that the hadron has exchanged a pomeron with the rest of the particles involved and that the pomeron has participated in the hard interaction .
the possibility of such interactions was proposed by ingelman and schlein @xcite on the grounds that the entity exchanged in elastic scattering , called the pomeron , must be made of quarks and gluons , which , being pointlike , can participate in hard interactions .
the theoretical ideas and formulas involved are elaborated in some detail in ref .
the predicted phenomenon was seen in jet production in hadron collisions by the ua8 collaboration @xcite . as discussed in our previous work @xcite
, the ingelman - schlein model can be thought of as involving a `` diffractive parton distribution function , '' which is the subject of this paper .
the idea is that this function , @xmath1 represents , in a hadron of type @xmath2 , the probability per unit @xmath3 to find a parton of type @xmath4 carrying momentum fraction @xmath5 , while leaving hadron @xmath2 intact except for the momentum transfer characterized by parameters @xmath6 . here
@xmath7 is the invariant momentum transfer @xmath8 while @xmath9 is the fraction of its original longitudinal momentum lost by the hadron .
the parameter @xmath10 is the factorization scale , roughly , the resolution of the parton probe . a function expressing the same physics as the diffractive parton distribution ( [ fdiff ] ) has been proposed by veneziano and trentadue @xcite under the name of `` fracture function . ''
the details are a little different , as we will explain in sec .
[ diffdis ] .
the original paper of ingelman and schlein did not mention the function ( [ fdiff ] ) but instead introduced a related function , the `` distribution of partons in the pomeron . ''
our purpose in this paper is , first of all , to relate these various functions and the ideas behind them to one another and to comment on the likely validity of the formulas that express cross sections in terms of these functions .
we give operator definitions for the diffractive parton distribution functions and discuss the evolution equation that they obey .
we briefly review the expected behavior of @xmath11 for small @xmath12 , where @xmath13 .
then we use a perturbative calculation to explore how these functions behave for small values of @xmath14 . in the ingelman - schlein language ,
our results favor a rather `` hard '' distribution of partons in the pomeron .
we also present , in an appendix , a calculation of @xmath15 in a simple model .
we conclude with some observations on the experimental consequences of the theory .
in a deeply inelastic scattering reaction , a hadron @xmath2 with momentum @xmath16 is struck by a far off shell photon with momentum @xmath17 .
it is convenient to use momentum components @xmath18 , where @xmath19 , and where we denote transverse components of vectors by boldface .
we work in the brick wall frame , in which @xmath20 , { \bf 0})$ ] and @xmath21 .
one measures the standard hard scattering variables @xmath22 and @xmath23 .
in some deeply inelastic scattering events there will be in the final state a diffractively scattered hadron @xmath24 with momentum @xmath25p_{\!a}^+,{{\bf p}_{\!a^\prime}^{2 } + m_{\!a}^2\over 2[1-x_\pom]p_{\!a}^+ } , { \bf p}_{\!a^\prime}\right ) \label{paprime } % \ ] ] as in fig .
( [ diffdis ] ) .
the hadron has lost a fraction @xmath9 of its plus momentum and has gained transverse momentum @xmath26 .
the invariant momentum transfer from the proton , @xmath27 , is @xmath28 the events in which we are interested have small @xmath7 .
one expects @xmath29 to be typical .
we also suppose that @xmath9 is rather small .
one expects pomeron physics to be dominant for @xmath30 .
having found such events , one can construct the contribution to @xmath31 from final states containing a diffractively scattered hadron with variables @xmath7 and @xmath9 : @xmath32 .
( diffdis.ps width 10 cm ) 0.5 cm the model of ingelman and schlein @xcite as applied to deeply inelastic scattering is simple to state .
we begin with the usual factorization theorem for the structure function @xmath31 : @xmath33 here @xmath34 is the distribution of partons of kind @xmath4 in hadron @xmath2 as a function of momentum fraction @xmath5 , as measured at a factorization scale @xmath10 , while @xmath35 is the structure function for deeply inelastic scattering on parton @xmath4 . if , for simplicity , we ignore @xmath36 exchange , then @xmath37 is @xmath38 thus @xmath31 is rather trivially related to the parton distribution functions at the born level ; nevertheless , conceptually the distinction between @xmath39 and @xmath34 is quite important . as in our previous paper @xcite ,
we break the analysis into two stages . in the first stage , we hypothesize that the diffractive structure function @xmath40 can be written in terms of a diffractive parton distribution , eq .
( [ fdiff ] ) : @xmath41 in the second stage , we hypothesize that @xmath42 has a particular form : @xmath43 here @xmath44 is the pomeron coupling to hadron a and @xmath45 is the pomeron trajectory .
we distinguish the `` regge factorization '' of eq .
( [ pomdist ] ) from the `` diffractive factorization '' of eq .
( [ factor ] ) . in eq .
( [ pomdist ] ) we adopt standard conventions such that the proton - proton elastic scattering amplitude is @xmath46 then the elastic scattering cross section is @xmath47 while the total proton - proton cross section is @xmath48\ , s^{\alpha(0)-1}\ , .
% \ ] ] the normalization factor @xmath49 in eq .
( [ pomdist ] ) is quite arbitrary . here
, we have adopted the convention of donnachie and landshoff @xcite .
the function @xmath50 thus defined is the `` distribution of partons in the pomeron . '' in writing eq .
( [ pomdist ] ) , one thinks of the pomeron as a continuation in the angular momentum plane of a set of hadron states .
since hadrons contain partons , the pomeron should also .
thus one has in eq .
( [ pomdist ] ) the standard factors describing the coupling of the pomeron to hadron a , together with a distribution of partons in the pomeron @xcite .
inserting eq .
( [ pomdist ] ) into ( [ factor ] ) , one obtains the model of ingelman and schlein , applied to the case of deeply inelastic scattering : @xmath51 where @xmath52 .
we offer here a word of caution . both the structure of eq .
( [ is ] ) and the language `` distribution of partons in the pomeron '' suggest that the hadron emits a pomeron some long time before the hard interaction and that the pomeron then splits into partons , one of which participates in the hard interaction .
this interpretation is , however , not required by eq .
( [ is ] ) and is surely quite misleading . in a diagrammatic interpretation of pomeron exchange @xcite ,
the exchanged quanta have small plus and minus components of momentum .
thus the exchange takes place over a long interval @xmath53 in space - time .
it begins long before the hard interaction and ends long afterwards .
our diagrammatic analysis in secs .
[ tauglue ] and [ tauquark ] will provide an illustration of this picture .
we see that eq .
( [ factor ] ) , can be regarded as a version of the ingelman - schlein model ( [ is ] ) that is more parsimonious in its assumptions .
( [ factor ] ) says only that factorization still applies when hadron @xmath2 is diffractively scattered .
the ingelman - schlein model ( [ is ] ) assumes that regge phenomenology is applicable and , with the aid of this assumption , has more predictive power . in this paper , we concentrate on the case in which hadron @xmath24 is the same kind of hadron as hadron @xmath2 , so that vacuum quantum numbers are exchanged , and we consider @xmath9 to be small enough so that pomeron exchange dominates .
one should keep in mind , however , that eq .
( [ factor ] ) admits generalizations to cases where @xmath54 and where @xmath9 is not at all small .
one can also generalize eq .
( [ is ] ) to @xmath54 , but then @xmath9 should be fairly small in order that just one or two regge exchanges dominate . the diffractive factorization equation ( [ factor ] ) , or rather a very closely related equation , has been introduced by veneziano and trentadue @xcite .
these authors call the analogue of @xmath55 a `` fracture function . ''
stated precisely , a fracture function is @xmath56 by integrating over @xmath7 , veneziano and trentadue eliminate a variable that is perhaps of secondary importance .
however , there is some advantage to _ not _ integrating over @xmath7 .
we are interested in the physics of diffraction , which occurs in the small @xmath7 region .
if we integrate over @xmath7 , then we are forced to consider also the large @xmath7 region , in which the hadron @xmath24 is to be thought of not as the original hadron appearing in a scattered form but as a random hadron in a high @xmath57 jet produced in the hard interaction . taking this possibility into account
leads to certain complications in the formulas . in the following sections , we analyze the diffractive parton distributions . according to eq .
( [ factor ] ) , the measured quantity @xmath58 is approximately the sum of diffractive quark distributions weighted by the square of the quark charges .
there are higher order corrections to this relation , some involving the diffractive gluon distribution .
thus these distributions are rather directly related to experiment .
the reader may wonder why we concentrate on the theoretical diffractive parton distributions rather than on the physical quantity @xmath58 .
the reasons are the same as in ordinary hard scattering : 1 ) the diffractive parton distributions are process independent and 2 ) the factorization ( [ factor ] ) allows one to include perturbative corrections to the hard scattering .
the reader may also wonder why we do nt frame the analysis in terms of the distribution of partons in the pomeron .
our excuse is ignorance .
we do nt know how to relate the regge factorization in eq .
( [ pomdist ] ) to quantum field theory .
in this section we give an operator definition of the diffractive parton distribution .
we write the ordinary distribution of a quark of type @xmath59 in a hadron of type @xmath2 in terms of field operators @xmath60 evaluated at @xmath61 , @xmath62 @xcite : @xmath63 similarly , the ordinary distribution of a gluon in a proton is written as @xmath64 the proton state @xmath65 has spin @xmath66 and momentum @xmath67},{\bf 0})$ ] .
we average over the spin .
our states are normalized to @xmath68 the field @xmath69 is the quark field operator modified by multiplication by an exponential of a line integral of the vector potential : @xmath70 \psi_j(0,y^-,{\bf 0})\ , .
\label{tildepsi } % \ ] ] likewise @xmath71 is defined by @xmath72_{ab } f_b(0,y^-,{\bf 0})^{\mu\nu}\ , .
\label{tildef } % \ ] ] the @xmath73 denotes path ordering of the exponential .
the matrices @xmath74 in eq .
( [ tildepsi ] ) are the generators of the @xmath75 representation of su(3 ) , while in eq .
( [ tildef ] ) they are the generators of the @xmath76 representation .
these operator products are ultraviolet divergent , and are renormalized at scale @xmath10 using the @xmath77 prescription , as described in @xcite .
the motivation for these definitions is that in qcd canonically quantized on null surfaces @xmath78 using @xmath79 gauge , the operators measure the probability to find a quark and gluon respectively carrying plus component of momentum equal to @xmath80 .
the line integrals of the color potential restore gauge invariance
. then @xmath77 renormalization removes divergences .
the line integrals of the color potential have a physical interpretation . whenever a parton is measured by a short distance probe , the color carried by that parton has to go somewhere .
for instance , in deeply inelastic scattering , the color is carried away by the recoiling struck quark . in the definition of the parton distribution function ,
the recoil color flow is idealized as an infinitely narrow jet moving with the speed of light along the path @xmath81 with @xmath82 .
any gluons from the color field of the hadron can couple to this idealized color source .
consider now the diffractive distribution of a quark in a proton .
the operator is the same as in eq .
( [ quarkdist ] ) , but the proton is required to appear in the final state carrying momentum @xmath83 : @xmath84 we sum over the spin @xmath85 of the final state proton and over the states @xmath86 of any other particles that may accompany it .
similarly , the diffractive distribution of gluons in a hadron is @xmath87 the green function @xmath88 for a parton of type @xmath4 can , at least in principle , be computed from feynman diagrams together with the bethe - salpeter wave functions for the bound states .
we will want to change variables to @xmath9 and @xmath7 as defined in the previous section .
using eqs .
( [ paprime ] ) and ( [ t ] ) , we obtain @xmath89 integrating over the azimuthal angle @xmath90 , we have @xmath91 where @xmath92 for quarks is given in eq .
( [ fdiff1 ] ) and @xmath93 for gluons is given in eq .
( [ fdiff2 ] ) .
as mentioned in the previous section , the diffractive parton distributions are ultraviolet divergent and require renormalization .
it is convenient to perform the renormalization using the @xmath77 prescription , as discussed in @xcite .
this introduces a renormalization scale @xmath10 into the functions . in applications ,
one sets @xmath10 to be the same order of magnitude as the hard scale of the physical process .
the renormalization involves ultraviolet divergent subgraphs , such as that shown in fig .
[ evolution](a ) .
subgraphs with more than two external parton legs carrying physical polarization , such as that shown in fig .
[ evolution](b ) , do not have an overall divergence .
thus the divergent subgraphs are the same as for the ordinary parton distributions .
we conclude that the renormalization group equation for the diffractive parton distributions is @xmath94 with the same dglap kernel @xcite , @xmath95 , as one uses for the evolution of ordinary parton distribution functions .
( evolution.ps width 10 cm ) 0.5 cm .
[ evolution ] if , following veneziano and trentadue , we integrate over @xmath7 , then the large @xmath7 integration region introduces new ultraviolet divergences and the renormalization group equation is modified @xcite . in this paper , we choose to restrict integrations over @xmath7 to the small @xmath7 region .
in sec . [ diffdis ] , we presented the hypothesis of diffractive factorization for diffractive deeply inelastic scattering , as represented by eq .
( [ factor ] ) : @xmath96 this is an example of the more general hypothesis of factorization for other kinds of diffractive hard scattering .
another example is diffractive jet production .
consider , for example , the inclusive cross section for the production of two jets in a high energy collisions of two hadrons , @xmath2 and @xmath97 .
( at hera , this would be @xmath98 where hadron @xmath97 is the the hadronic or `` resolved '' part of the photon . )
let the initial hadron @xmath2 have momentum @xmath99 while hadron @xmath97 enters the scattering with momentum @xmath100 we specify the two jets by variables @xmath101 , @xmath102 , and @xmath103 , given in terms of the four momenta @xmath104 and @xmath105 of jets @xmath106 and @xmath107 by @xmath108 if we add the requirement that hadron @xmath2 emerge scattered with scattering parameters @xmath6 , then we have diffractive jet production .
the corresponding hypothesis of diffractive factorization for this cross section is @xmath109 of course , eqs .
( [ factorencore ] ) and ( [ factorjets ] ) are approximations , as indicated by the @xmath110 signs .
we understand the hypothesis of diffractive factorization to mean that the corrections to these relations are suppressed by a power of @xmath111 or @xmath112 , where @xmath113 represents the momentum scale of soft hadronic interactions and @xmath114 or @xmath101 is the scale of the hard interaction .
this general diffractive factorization hypothesis was put forward by veneziano and trentadue in their paper @xcite on fracture functions . the ingelman - schlein model @xcite demands diffractive factorization plus the regge structure of the diffractive parton distributions .
thus , in a strict interpretation , the validity of the ingelman - schlein model logically implies the validity of diffractive factorization . on the other hand , one might interpret the ingelman - schlein model as being valid if corrections to it , while not vanishing in the limit of large @xmath114 , were nevertheless _ numerically _ small .
thus , for instance , the authors of ref .
@xcite speculated that the factorization inherent in the ingelman - schlein model was not likely to be exact up to @xmath111 corrections but might have the same status as regge factorization , which has proven to be a useful approximation even if it is not exact .
the hypothesis of diffractive factorization appears to us to be correct in the case of deeply inelastic scattering .
a detailed proof of this statement is beyond the scope of this paper .
however , we can briefly sketch how such a proof would go , following the ideas of refs .
the singularities of the cut feynman graphs for diffractive deeply inelastic scattering are such that the leading integration regions involve 1 ) a beam jet in the direction of the initial hadron @xmath2 ( which includes the final state diffracted hadron @xmath24 ) , 2 ) a hard interaction , 3 ) one or more final state jets that are _ not _ in the direction of hadron @xmath2 , and 4 ) possible soft gluons ( sometimes with soft quark loops ) that may communicate between the beam jet and the final state jets .
factorization would be more or less kinematic were it not for the possibility of the soft gluons linking the beam jet with the final state jets .
one must use gauge invariance to show that the soft gluons do nt really `` see '' the details of the final state jets , so that the connections to the final state jets can be replaced by connections to the idealized jet that is embodied in the line integral of the field operator @xmath115 in the definitions of the diffractive parton distribution functions , eqs .
( [ quarkdist],[gluedist],[tildepsi],[tildef ] ) .
what of diffractive factorization for processes with two hadrons in the initial state , such as @xmath116 or @xmath117 ? here
the proof of _ ordinary _ factorization is much more delicate .
the problem is that low momentum gluons can communicate between the partons of the two beam jets .
this can happen even before the hard scattering takes place , as in fig .
[ diffjetbad ] .
when one looks for the hard process inclusively , such effects cancel @xcite .
but the cancellation requires a sum over final states .
collins , frankfurt , and strikman @xcite have argued that the demand that the final state include a diffractively scattered hadron destroys the factorization . in ref .
@xcite , we looked at this problem in the context of a simple perturbative model .
we found that the diffractive factorization hypothesis is , indeed , not valid .
new terms proportional to @xmath118 appear in eq .
( [ factorjets ] ) .
( presumably in a more general model one will also have factorization violating terms that are not proportional to @xmath118 but are singular as @xmath119 . )
these new terms have an interesting structure that can be investigated experimentally at hera .
( diffjetbad.ps width 10 cm ) 0.5 cm
the diffractive parton distribution @xmath121 is essentially a long distance object , which is not amenable to calculation using perturbative methods .
however , regge phenomenology provides an expectation for the behavior of @xmath122 when the parameter @xmath123 is small compared to 1 . in the case of the ordinary parton distributions , this expectation is @xmath124 where @xmath125 is the pomeron intercept .
that is , one expected @xmath126 for small @xmath127 .
the simplest version of this expectation is that @xmath125 is the same as in soft pomeron physics , @xmath128 @xcite .
however , this expectation is hard to reconcile with evolution : if it holds at some rather small value of @xmath129 , then @xmath31 should be steeper at larger values of @xmath129 @xcite .
a steeper dependence was also expected on the basis of the perturbative version of the pomeron analyzed at leading log level @xcite .
indeed , a steeper dependence , with @xmath130 , is found in experiments at hera @xcite .
the analogous expectation @xcite for the @xmath12 dependence of diffractive parton distributions at fixed @xmath9 is @xmath131 in this case , the constant is proportional to the triple pomeron coupling used in regge physics .
presumably , @xmath125 here should not be the true pomeron intercept @xmath132 but should be the larger value , @xmath133 , found in the hera experiments .
then , of course , it is debatable whether the triple pomeron coupling used here should be the same as that found in soft regge physics .
uncertainty over the precise values , however , should not obscure the simple prediction that the diffractive @xmath31 should behave at small @xmath12 like @xmath134 with @xmath135 .
the small @xmath12 behavior of @xmath40 is analyzed in refs .
the diffractive parton distribution @xmath121 , like the ordinary parton distribution , is essentially not calculable using perturbative methods .
recall , however , that it is possible to derive `` constituent counting rules '' that give predictions for ordinary parton distributions @xmath137 in the limit @xmath138 for not too large values of the scale parameter @xmath10 @xcite . in the same spirit , we consider in this section and the next the diffractive parton distributions in the limit @xmath136 where @xmath139 .
our analysis is similar to that if ref .
the authors of that paper concluded , with certain caveats , that the diffractive gluon distribution should behave like @xmath140 as @xmath136 .
our reanalysis suggests a behavior between @xmath140 and @xmath141 , depending on how certain nonperturbative issues are resolved .
for the diffractive quark distribution ( not treated in ref .
@xcite ) , our analysis in sec .
[ tauquark ] suggests a behavior between @xmath142 and @xmath140 .
this analysis involves both hard and soft subprocesses , put together in a manner that is not under solid theoretical control .
thus there is not a clearly correct final answer , as far as we can see .
what we do here is to provide some calculational results that can help restrict the range of answers and give some basis for the reader s informed judgment .
. then we examine @xmath144 in the limit @xmath145 , after first taking @xmath146 so as to separate out pomeron exchange from other regge pole exchanges .
we consider the diffractive gluon distribution first ( _ a = g _ ) .
later , we supply the modifications needed for the diffractive quark distribution . throughout this section , except
as specifically noted , we consider feynman graphs in null plane gauge @xmath79 .
our analysis is based on a model that consists of a selected set of feynman graphs , with the internal loop momenta integrated over a selected integration region . within this model
there is a hard subgraph in which all internal propagators are far off - shell and a soft subgraph in which the propagators are near to being on - shell .
we evaluate the hard subgraph at lowest order in perturbation theory .
we offer a perturbatively based conjecture about the behavior of the soft subgraph in the regge limit @xmath147 .
the simple graphs considered here do not include the graphs that contribute to qcd evolution of the parton distributions .
thus we imagine that the analysis applies to the diffractive parton distribution at a starting scale @xmath148 that is not too large ( say , 2 gev ) .
standard parton evolution starting at this scale will soften the distributions . according to eq .
( [ fdiff2 ] ) , the diffractive gluon distribution is the square of the matrix element of an operator that destroys a gluon with longitudinal momentum fraction @xmath149 , where the matrix element is taken between the initial proton state and a final state that includes the scattered hadron plus anything else .
since a color octet gluon is destroyed , while the initial and final hadrons are color singlets , the final state must include at least one gluon .
we consider here the minimal model in which the final state includes precisely one gluon .
then this gluon carries a very small momentum fraction @xmath150 .
call the momentum of the final state gluon @xmath17 , as depicted in fig .
[ structurefig ] .
since this gluon is on - shell , we have @xmath151 before proceeding , we pause for a technical point . in general in this section we use the null plane gauge @xmath152 , but for the final state gluon this is not convenient .
the polarization vectors @xmath153 for transverse polarization in the @xmath154 direction have minus components that grow like @xmath155 in the limit @xmath145 : @xmath156 we avoid this singular behavior by changing the polarization vectors for this gluon to @xmath157 gauge .
the difference @xmath158 between the old polarization vector and the new is proportional to @xmath17 , so that changing polarization vectors has no effect after we sum over a gauge invariant set of graphs . in @xmath159 gauge ,
@xmath160 , @xmath161 and @xmath162 thus @xmath153 is predominantly transverse , with a plus component that vanishes as @xmath145
. some vocabulary will be helpful for discussing the physics of time and momentum scales in this problem .
there are three relevant longitudinal momentum scales .
we call partons with @xmath163 fast partons . for instance
, the valence quarks in a hadron are typically fast partons .
we call partons with @xmath164 slow partons . finally , we call partons with @xmath165 very slow partons .
the final state gluon is such a parton .
the gluon cloud surrounding any hadron contains gluons at any momentum fraction .
consider a gluon with momentum fraction @xmath127 and transverse momentum of order @xmath113 , where @xmath166 gives the scale of a typical hadronic mass or transverse momentum .
the contribution of such a gluon to the null plane energy @xmath167 of an intermediate state is @xmath168 , which is the minus momentum of a free gluon with transverse momentum @xmath169 and plus momentum @xmath170 .
this `` kinetic '' minus momentum is of order @xmath171 .
thus a gluon of the type that we are calling a slow gluon has a kinetic null plane energy @xmath172 and survives for a typical null plane time @xmath173 .
notice that the final state gluon has a large minus momentum , @xmath174 , at least as long as its transverse momentum @xmath175 is not too small .
now @xmath175 is not observed ; we are to square the matrix element and integrate over @xmath175 .
we can not say anything about the region of very small @xmath176 , but we can analyze the contribution to the integral from the region @xmath177 defined by @xmath178 we will consider the contribution to @xmath179 from the region @xmath177 with the hope that the contribution from the complementary region @xmath180 , is not large enough to overwhelm or , worse , to cancel the contribution from region @xmath177 .
we will also evaluate the contribution from the smaller integration region in which the transverse momentum @xmath181 is large : @xmath182 .
since this region is described by rather standard short distance dynamics , its contribution should provide a lower bound on the true result .
( again , we assume that this contribution is not canceled by some long distance contribution . ) for @xmath183 the minus momentum @xmath184 of the final state gluon is large compared to the kinetic minus momentum @xmath185 of a typical fast parton and is also large compared to the kinetic minus momentum @xmath186 of a typical slow parton .
this large minus momentum flows through the graph and is carried out of the graph by the detected gluon .
thus the parton detection with @xmath145 creates a hard process that happens on a null plane time scale @xmath187 that is short compared to the typical time scale for interactions within the proton or its cloud of slow gluons .
we base our analysis on this observation , using low order perturbation theory for the nearly instantaneous interaction that probes the parton distribution .
( structure.ps width 8 cm ) 0.5 cm [ structurefig ] three graphs that involve the lowest order hard interaction are shown in fig .
[ hardfig ] . in these graphs
, the large minus momentum @xmath188 flows through only one or two propagators , as indicated by the heavy lines .
these propagators are far off shell .
we refer to this part of the graph as the hard subgraph .
the rest of the graph is the soft subgraph .
( hard.ps width 15 cm ) 0.5 cm [ hardfig ] there are also two graphs of the same order that involve connecting the @xmath17 and @xmath189 gluons to a quark line , as depicted in fig . [ slowquarkfig ] .
these graphs require a separate discussion , which we omit in this paper on the grounds that low momentum gluons are likely to be dominant in the pomeron compared to low momentum quarks .
( slowquark.ps width 10 cm ) 0.5 cm in fig .
[ hardfig ] , we denote by @xmath190 and @xmath191 , respectively , the momenta of the gluon leaving the soft subgraph and entering it after the hard interaction .
let the momentum fraction of the first gluon be @xmath192
. then momentum conservation fixes @xmath193 .
we integrate over @xmath194 .
it is convenient to distinguish between the two identical gluons by requiring that @xmath195 .
that is , @xmath196 . in principle
, the @xmath194 integration runs to @xmath197 , but the reader can verify that the region @xmath198 is not important . we take it as a working assumption , to be verified later , that the region @xmath199 is also not important . thus the important integration region is @xmath200 .
that is , the gluons that couple the hard subgraph to the soft subgraph are typical `` slow '' gluons .
we integrate over the transverse momentum @xmath201 , setting @xmath202 .
we also integrate over @xmath203 , setting @xmath204 .
we suppose that the hadron wave functions fix @xmath201 and @xmath203 to be no larger than the ordinary size for slow gluons , @xmath205 and @xmath206 .
( contributions in which @xmath190 is part of a hard virtual loop are more properly considered to be part of a higher order correction to the hard subgraph . )
taking the limit @xmath145 , we find that if we consider the hard subgraph to be a function of @xmath190 , @xmath191 , and @xmath17 , then it is independent of @xmath203 , @xmath207 , @xmath201 and @xmath208 and is also also independent of @xmath9 .
in addition , the dominant polarizations for the gluons entering the hard subgraph from the soft subgraph are transverse ; in a shorthand notation , @xmath209^{\mu\nu}[{\rm soft}]_{\mu\nu } \approx \sum_{i , j = \{1,2\}}[{\rm hard}]^{ij}[{\rm soft}]_{ij}. \label{hardsoftindices } % \ ] ] that the hard subgraph is independent of @xmath203 and @xmath207 is not surprising , since @xmath210 and @xmath211 . that it is independent of @xmath9 follows simply from its invariance under boosts in the longitudinal direction .
it is not , however , obvious that the hard subgraph is independent of @xmath201 and @xmath208 and that only transverse polarizations are important .
these results follow from an argument that we describe in appendix [ appendix : hardstructure ] .
the picture that emerges is one in which the soft gluon cloud surrounding the hadron is probed by an interaction that is effectively local in null plane time , @xmath212 and in transverse position @xmath213 .
we take advantage of these results by setting @xmath214 and @xmath215 in the hard subgraph and restricting the polarization sum to transverse polarizations .
then the diffractive distribution function depends on the integral over @xmath203 and @xmath201 of the soft subgraph .
we write this structure as depicted in fig .
[ diffgluefig ] , @xmath216 here the first line represents the soft subgraphs while the second line represents the integral of the hard subgraphs .
we discuss these factors below .
the function @xmath217 is the amplitude for the proton to emit a gluon with transverse polarization @xmath218 and momentum fraction @xmath219 then absorb a gluon with transverse polarization @xmath154 and momentum fraction @xmath220 .
@xmath221 there is a summation over the color @xmath4 of the gluon operators ; in @xmath222 we average over these colors .
the factor @xmath223 arises from changing the integration in eq .
( [ diffint ] ) from @xmath224 to @xmath225 .
we now turn to the hard interaction .
we begin by considering the probed gluon , which carries momentum @xmath189 . according to eqs .
( [ fdifftog ] ) and ( [ fdiff2 ] ) , the operator that probes the gluon distribution is @xmath226 .
( only @xmath227 contributes in eq .
( [ fdiff2 ] ) . ) in @xmath152 gauge , this is simply @xmath228 .
( we have changed the polarization vector @xmath229 for the final state gluon from @xmath230 , so this gluon could couple to the @xmath231 operator in @xmath232 . however , this graph is not allowed when the two exchanged gluons are in a color singlet state . ) the @xmath233 becomes a @xmath234 that we absorb into the normalization .
this leaves a propagator for the probed gluon , @xmath235 thus each probe operator gives a @xmath236 factor and leaves the amputated hard interaction graph dotted into a polarization vector for a gluon of momentum @xmath189 , polarization @xmath113 , in @xmath152 gauge .
we call the amputated hard interaction graph @xmath237 .
the indices @xmath113 and @xmath238 are the transverse polarizations of the probed and final state gluons respectively ; @xmath4 and @xmath239 are their colors . recall that the final state gluon polarization is represented by a polarization vector @xmath240 is , by our convention , in @xmath157 gauge , as in eq .
( [ aminusgauge ] ) . in @xmath222 ,
the exchanged gluons are approximated as having zero transverse and minus momenta ; one integrates over these momenta in the definition of @xmath2 .
the exchanged gluons have transverse polarizations @xmath218 and @xmath154 respectively . in the @xmath145 limit ,
the factor @xmath241 in the denominator is @xmath242 thus the probed gluon is far off shell , @xmath243 , as long as @xmath176 is in the integration region @xmath177 , @xmath244 .
similarly , the virtual lines internal to @xmath222 are far off shell and have a simple form when @xmath183 .
this is so even though the region @xmath177 includes small transverse momenta , @xmath245 .
indeed , the small transverse momenta near the boundary of @xmath246 dominate the integral in eq .
( [ diffint ] ) , as we will see .
next , we examine the soft subgraphs and then the hard subgraphs in some detail .
( diffglue.ps width 15 cm ) 0.5 cm in this subsection , we investigate the @xmath9 dependence of @xmath247 in the limit @xmath248 . comparing the expected regge form ( [ pomdist ] ) of @xmath249 with eq .
( [ diffint ] ) , we see that the expected regge form of @xmath250 is @xmath251 the question is , what is the pomeron trajectory @xmath45 within the context of the analysis used in this paper ? the operators eq . ( [ aij ] ) are in @xmath79 gauge , but the amplitude can be put in a gauge invariant form by reexpressing it in terms of the gluon field operators @xmath252 defined in eq .
( [ tildef ] ) .
we simply replace @xmath253 and make a similar replacement for @xmath254 .
the @xmath255 choice here is of some significance
. we will discuss it in the following subsection .
the resulting form for @xmath250 is @xmath256 since the propagators in the soft subgraph are @xmath257 far off - shell , one is not really justified to use perturbation theory to investigate @xmath247 . nevertheless , perturbation theory is suggestive .
consider eq .
( [ aijmod ] ) using feynman gauge .
suppose that the gluon annihilated by the operator @xmath258 connects to an on - shell quark carrying no transverse momentum and plus momentum @xmath259 , with @xmath260 of order 1 .
that is , the gluon couples to a `` fast '' quark .
the relevant factor is @xmath261 where @xmath262 . in the limit @xmath263 , this is independent of @xmath9 .
similarly , the coupling of the operator @xmath264 to a fast quark gives no @xmath9 dependence .
thus perturbation theory suggests that the operator matrix element in eq .
( [ aijmod ] ) is independent of @xmath9 for small @xmath9 .
considering that there is a factor @xmath265 in eq .
( [ aijmod ] ) , we find that the pomeron trajectory appearing in eq .
( [ zdependenceofa ] ) is @xmath266 in this most naive analysis of the soft subgraph .
this is close to the pomeron trajectory observed in nature , but of course this most naive analysis is too naive , and we expect that soft interactions among the gluons modify the result .
what we see here is that the picture of the pomeron as two gluon exchange , which often gives results that are surprisingly good considering the simplicity of the picture @xcite , works rather well also in this context . the analysis of the preceding subsection has revealed a @xmath267 singularity within the integration domain @xmath268 of the momentum fraction @xmath194 .
what is the nature of this singularity ?
if we stick to @xmath79 gauge , one can check that it arises from the @xmath269 singularity in the gluon propagator , @xmath270 .
% \ ] ] this singularity is usually interpreted with a principle value prescription , but there is no compelling reason for this choice .
in fact , @xmath79 gauge is not an effective tool for examining the nature of the @xmath267 singularity , since in this gauge the singularity is a gauge artifact .
thus we choose to examine this question in feynman gauge , in the style of ref .
( arguably , it would be better to do the whole problem in feynman gauge , but this leads to its own complications . )
( singularity.ps width 10 cm ) 0.5 cm we consider the graph shown in fig .
[ singularityfig ] in feynman gauge but with a transverse polarization chosen for the gluon carrying momentum @xmath190 .
it is helpful to choose a frame in which @xmath271 .
then @xmath272\ll m \nonumber\\ q^+ = \tau x_\pom p_{\!a}^+ \ll m & & \quad q^- = { \bf q}^2/[2 \tau x_\pom p_{\!a}^+ ] \gg m. % \end{aligned}\ ] ] the graph contains the structure @xmath273 the denominator @xmath274 has the form @xmath275 now @xmath188 is large and positive while @xmath207 is small in the dominant integration region . thus the denominator has the approximate form @xmath276 } \label{sdenom } % \ ] ] where the dots indicate small terms .
this denominator contains the only important dependence on @xmath277 as long as @xmath278 in @xmath279 .
notice that there is a singularity very near to @xmath280 , but that , since there are no other singularities nearby , we can deform the integration contour away from this singularity .
thus we deform the @xmath277 contour into the upper half complex @xmath277 plane . on the deformed contour , we have @xmath281 or @xmath200 .
now we examine the numerators .
the largest component of the quark current @xmath282 is @xmath283 , while the largest component of the current @xmath284 of the final state gluon is @xmath285 .
thus @xmath286 .
thus the dominant term in @xmath287 is obtained by replacing @xmath288 the factor @xmath289 is approximately @xmath290 as long as @xmath277 is not small , and we know that @xmath277 _ is _ not small since we have deformed the integration contour so that it does not approach @xmath291 .
the next step is to restore the @xmath277 integration contour to the real axis , taking care not to cross any singularities .
this means that we should move the @xmath269 singularity in eq .
( [ replacement0 ] ) infinitesimally into the lower half @xmath277 plane , so that our replacement becomes @xmath292 this is in keeping with the usual notation in which integration contours are along the real axis , with poles infinitesimally displaced from the integration contour .
the replacement ( [ replacement ] ) gives the dominant contribution to our graph .
however , if we attach the gluon carrying momentum @xmath191 everywhere in the hard subgraph and sum the leading terms obtained with this replacement , we will get zero because of the ward identities obeyed by the hard graph and because the two gluons carrying momenta @xmath190 and @xmath191 together form a color singlet .
( the relevant identities are discussed in the appendix of ref .
@xcite ) .
we have thus encountered the _ bte noir _ of feynman gauge : the leading contributions graph by graph come from unphysical polarizations that cancel when one sums over graphs .
what we need is the subleading contribution .
that is easy .
we replace @xmath293 where @xmath294 now we throw away the first term in ( [ rereplacement ] ) , since it will cancel , and keep the remainder .
this gives the structure in eq .
( [ ajreplacement ] ) , but now with a prescription for the @xmath269 singularity determined on physical grounds .
we now turn to the hard interaction function @xmath222 .
it is a simple matter to evaluate this function .
we find @xmath295 .
\label{calmresult } % \end{aligned}\ ] ] thus @xmath296 as @xmath145 .
the @xmath255 prescription in the @xmath267 factor arises from eq .
( [ sdenom ] ) .
it matches the @xmath255 prescription in the soft function @xmath297 , eq .
( [ aijmod ] ) , so that , after a contour deformation , @xmath194 is never much smaller than 1 . inserting eq .
( [ calmresult ] ) and @xmath298 into eq .
( [ diffint ] ) , we obtain @xmath299 here @xmath300 is a rational function of @xmath194 and @xmath301 that is not of particular interest .
recall that we integrate @xmath175 over the region @xmath177 defined by @xmath244 ( eq . ( [ calr ] ) ) , where @xmath113 is a typical hadronic mass or transverse momentum . here
we make the prescription more precise by integrating over @xmath302 where @xmath303 is any fixed mass such that @xmath304 . performing the integration in eq .
( [ diffint2 ] ) gives for the @xmath305 limit of the diffractive gluon distribution @xmath306 note that @xmath307 is independent of @xmath308 as @xmath145 : @xmath309 this may seem surprising , since the distribution of gluons in a typical hadron behaves like @xmath310 with a rather high power @xmath311 .
the leading behavior comes from the lower endpoint of the integration in eq .
( [ diffint2 ] ) .
thus it is sensitive to the cutoff chosen .
recall that we took @xmath302 because as long as @xmath312 , the internal lines in the hard subdiagram are far off - shell .
this appears to be the natural cutoff .
as soon as the internal lines of the subdiagram through which @xmath17 flows are _ not _ far off - shell , the the gluon that goes into the final state and the probed gluon can attach at different space - time points in the gluon cloud of the hadron .
then there is the opportunity for cancellation , as both gluons sample the color charge of the hadron as a whole and find that the hadron as a whole is a color singlet .
it is difficult to check this conjecture directly in a realistic model of nonperturbative hadron structure .
however we have checked in a very simple model where all the graphs can be included exactly .
this model is too simple to have the correct pomeron behavior , but we find that it does have @xmath313 behavior in the @xmath145 limit . the model is described in appendix [ appendix : model ] .
one might reasonably conjecture that a larger cutoff would be imposed by nonperturbative physics in a more realistic model .
for instance , if the gluon emitted into the final state effectively had a substantial mass @xmath314 , then @xmath188 would be @xmath315 .
this would induce an effective cutoff @xmath316 in eq .
( [ diffint2 ] ) .
then we would have obtained @xmath317 with @xmath318 .
this corresponds to the style of analysis of the constituent counting rules and gives the result @xmath140 found in ref .
presumably , the contribution from @xmath319 must be present and is not likely to be canceled , so that @xmath320 should not be smaller than @xmath140 at large @xmath12 .
that is , the power @xmath321 should not be larger than 1 .
we conclude that if the diffractive gluon distribution is parameterized as @xmath322 for @xmath136 at moderate values of the scale @xmath10 , say 2 gev , then @xmath323 the choice @xmath324 corresponds to an effectively massless final state gluon , while @xmath325 corresponds to an effective gluon mass .
in analogy with the gluon case , we write @xmath326 here the soft function @xmath327 is the amplitude for the proton to emit a gluon with polarization @xmath10 , momentum fraction @xmath328 and transverse momentum @xmath329 , then absorb a gluon with transverse polarization @xmath330 and momentum fraction @xmath220 and transverse momentum @xmath208 : @xmath331 the function @xmath332 is simply related to the operator matrix element @xmath250 , eq .
( [ aij ] ) , that appeared in our discussion of the diffractive gluon distribution : @xmath333 the function @xmath334 represents the amputated hard interaction graph . here
@xmath181 is the momentum of the antiquark that enters the final state and @xmath335 is the transverse momentum of the probed quark .
we have @xmath336 by momentum conservation .
the variables @xmath337 and @xmath338 are the helicities of the probed and final state quarks respectively ; @xmath339 and @xmath340 are their colors .
there is an important difference with the gluon case .
the quark has a mass @xmath341 .
thus the minus momentum of the on - shell quark entering the final state is @xmath342 then @xmath343 in the @xmath145 limit .
what counts here is the mass of the final state quark as it emerges from the hard interaction and propagates into the final state .
presumably the best model for @xmath341 in this role is the constituent quark mass ( @xmath344 ) , not the much smaller current quark mass .
this is a substantial mass , so that the condition that defined whether the virtual lines in @xmath222 are far off shell , @xmath345 is satisfied for any @xmath346 when @xmath308 is small .
thus we do not need to restrict the integration region . on the other hand ,
the region @xmath347 is not important in the integration .
we evaluate @xmath222 in the @xmath145 limit using null plane spin defined with the plus direction as special for the probed quark and defined with the minus direction as special for the final - state antiquark .
we find @xmath348 the @xmath349 are two component spinors @xmath350 and @xmath351 .
then we can write the @xmath352 matrix @xmath353 using transverse pauli spin matrices @xmath354 and @xmath355 . for transverse indices
@xmath356 , we find @xmath357 ( we hope that the pauli spin matrices @xmath358 will not be confused with the momentum fraction @xmath194 that occurs in this equation in the combinations @xmath267 and @xmath359 . ) for one transverse index and one plus index we find @xmath360 also @xmath361 , while @xmath362 does not give leading contributions as @xmath145 .
finally , we note that @xmath363 and @xmath364 are not needed because they multiply 0 in @xmath79 gauge .
the spin function @xmath365 is rather complicated , but we are concerned with only two of its properties .
first , if we think of @xmath353 in coordinate space as a function of the separation @xmath366 between the points where the two exchanged gluons attach , then @xmath353 is proportional to @xmath367 and to a linear combination of @xmath368 and @xmath369 .
thus the interaction is hard in the sense of being local in @xmath370 and @xmath371 .
second , and most important , @xmath353 is independent of @xmath308 .
we can now insert eqs .
( [ ksqquark ] ) and ( [ calmquark ] ) into eqs .
( [ diffintquark ] ) to obtain the @xmath308 dependence of the diffractive quark distribution : @xmath372 the crucial feature here is the factor of @xmath373 . we conclude that the constituent counting result for the diffractive distribution of quarks is @xmath374 however , suppose that we interpret the calculation of the previous section as saying that the diffractive distribution of gluons is proportional to @xmath141 for @xmath12 near 1 when the scale @xmath10 is not too large .
then the evolution equation for the diffractive parton distributions will give a quark distribution that behaves like @xmath375 when the scale @xmath10 is large enough that some gluon to quark evolution has occurred , but not so large that effective power @xmath321 in @xmath376 for the gluon distribution has evolved substantially from @xmath377 .
a signature of this phenomenon is that the diffractive quark distribution will be growing as @xmath10 increases at large @xmath12 , rather than shrinking .
perhaps this is seen in the data @xcite .
we close with some observations concerning the implications for hera physics of the discussion presented here .
we have discussed two kinds of factorization relevant to diffractive hard scattering .
both are experimentally testable .
we say that _ diffractive factorization _ holds if the cross section is a standard partonic hard scattering cross section convoluted with a diffractive parton distribution .
the diffractive parton distribution gives the distribution of partons in the hadron under the condition that the hadron is diffractively scattered .
( if there is a second hadron in the initial state and we do not demand that it be diffractively scattered , then the cross section should also contain a convolution with the ordinary parton distribution within that hadron . ) we say that _ regge factorization _ holds if the diffractive parton distribution is a product of @xmath378 times a pomeron - hadron coupling times a function @xmath379 that is interpreted as the distribution of partons `` in '' the pomeron .
these properties together constitute the ingelman - schlein model @xcite .
diffractive factorization is something that can be disproved for a given process by a counterexample at some fixed order of perturbation theory ( with wave functions for the hadronic states ) .
correspondingly , it could in principle be proved to hold at any fixed order of perturbation theory , which one would take as a strong indication that it holds beyond perturbation theory for that process .
diffractive factorization makes no statement about the regge structure of the nonperturbative factors .
regge factorization does make such a statement
. it would be interesting to study regge factorization from the point of view of the bfkl pomeron , perhaps extracting a model for the distribution of partons in the pomeron .
let us discuss first diffractive deeply inelastic scattering , which is simpler than diffractive hard scattering processes with two hadrons in the initial state . for diffractive deeply inelastic scattering ,
we argue that diffractive factorization is a consequence of perturbative qcd , although a detailed proof is beyond the scope of this paper . from the point of view of current theory ,
regge factorization for the diffractive parton distributions is a conjecture based on experience with soft diffractive scattering .
the hera experiments @xcite have now provided evidence concerning these issues .
it is a consequence of diffractive factorization that the diffractive structure function @xmath380 should exhibit approximate scaling as @xmath129 is increased with fixed @xmath9 and @xmath12 .
this is confirmed by the data . the dependence on @xmath9 in the form @xmath381 predicted by regge factorization , eq .
( [ is ] ) , is also confirmed by the data . for the future , it will be important to measure the diffractive parton distributions in as complete detail as possible , using charged and neutral current events , @xmath40 and @xmath382 , and probes for heavy flavors in the final state @xcite . especially crucial is the diffractive gluon distribution .
this can be measured using diffractive deeply inelastic scattering with high @xmath57 jets detected in the final state .
if we demand that we see two jets instead of the usual single struck quark jet , and if these two jets have a high transverse momentum relative to the direction determined by the sum of their momenta , then the hard process is of order @xmath383 instead of just @xmath384 .
for such a process , gluons participate as initial partons on the same footing as quarks .
since the diffractive quark distributions are already known , it should be possible to extract the diffractive gluon distributions .
our analysis suggests that the diffractive gluon distribution is quite hard , with a behavior between @xmath140 and @xmath141 for large @xmath13 at moderate values of the scaling parameter @xmath10 , say 2 gev .
the corresponding behavior of the diffractive quark distribution , which is quite directly measured in @xmath40 , is between @xmath142 and @xmath140 . here
the @xmath140 for quarks would arise if the diffractive gluon distribution is large and behaves like @xmath141 , so that the quarks at large @xmath12 are produced by @xmath385 .
given a complete set of diffractive parton distributions , it will be interesting to test the evolution equation ( [ apeqn ] ) .
let us now turn to hard processes with two hadrons in the initial state , as exemplified by @xmath386 at hera , where we look at the hadronic part of the photon . in this case ,
the factorization that holds in _ inclusive _ hard scattering is expected to break down in _ diffractive _ hard scattering , as shown by counterexamples at a fixed order of perturbation theory @xcite .
if one extracts diffractive parton distribution functions from deeply inelastic scattering and uses them to predict cross sections for @xmath386 , then the observed cross section should contain extra terms that do not match the prediction @xcite .
in particular , there should be extra contributions that correspond to the jets carrying almost all of the longitudinal momentum of the pomeron .
perhaps this corresponds to the `` superhard '' component seen in the ua8 experiment @xcite in @xmath387 . in summary ,
the hera results @xcite , together with the earlier ua8 results @xcite , have confirmed the basic features of the ingelman - schlein picture of diffractive deeply inelastic scattering .
the experiments have shown that diffractive scattering is related to exchanges of quanta that , when examined with a hard probe , appear to be the pointlike quarks and gluons of qcd .
much more remains to be done , but already we are challenged to connect the theory of pointlike gluons to the soft color dynamics that is presumably responsible for diffractive scattering .
we thank g. j. van oldenborgh for explicitly rechecking his algorithms in some of the sensitive low @xmath7 regions investigated in appendix [ appendix : model ] .
we thank j. c. collins , g. ingelman , j. bartels and many members of the h1 and zeus collaborations for helpful conversations .
in this appendix we investigate the structure of the hard subgraph for the diffractive distribution of gluons in hadron in the limit @xmath388 . recall that the hard subgraph is a function of momenta @xmath389 , and @xmath17 , where @xmath190 and @xmath191 are the momenta of the gluons exchanged with the soft subgraph and @xmath17
is the momentum of the gluon that goes into the final state . since @xmath390 we will consider the hard subgraph to be a function of @xmath181 and @xmath188 , replacing @xmath391 by @xmath392 . the momentum of the detected gluon is @xmath393 . in sec .
[ tauquark ] , we studied the hard subgraph in the limit that applies when @xmath388 .
this limit is really a dual limit .
we take @xmath394 but @xmath395 we also recall the definition ( [ calr ] ) of the integration region considered for @xmath181 , @xmath396 that is , @xmath397 . we combine this with the assumption that , in the effective integration region for @xmath201 , @xmath208 , @xmath203 and @xmath207 , these variables have values typical for slow gluons : @xmath398 then @xmath399 we claimed in sec .
[ decompose ] that in this limit , the hard subgraph is independent of the variables @xmath201 , @xmath208 , @xmath203 , and @xmath207 .
we also claimed that the transverse components of the hard subgraph dominate over other components in the limit considered , as in eq .
( [ hardsoftindices ] ) . in this appendix , we substantiate these claims .
we consider first the question of the independence on the variables @xmath201 , @xmath208 , @xmath203 , and @xmath207 , taking , for the moment , only the transverse components of the hard scattering subgraph .
this is the same as multiplying the hard subgraph by purely transverse polarization vectors for the two gluons exchanged with the soft subgraph .
we consider the hard amplitude @xmath222 , defined as in sec .
thus @xmath222 does not include the propagator for the detected gluon with momentum @xmath189 , but does include an @xmath230 gauge polarization vector for this gluon .
it also includes an @xmath400 gauge polarization vector for the gluon with momentum @xmath17 that enters the final state . as a matter of convenience
, we will analyze the first graph in fig .
[ hardfig ] .
essentially the same argument covers the second graph , while the third graph , with a four gluon interaction , is quite trivial .
the feynman rules give @xmath222 as a rational function of the components of @xmath389 , and @xmath17 .
we are interested in particular in the behavior of @xmath222 for small values of @xmath203 , @xmath207 , @xmath201 , @xmath208 and @xmath181 as specified in eqs .
( [ limitqplus ] ) and ( [ limitqminus ] ) . for this reason ,
we need to know if there are any factors of the small variables in the denominator of @xmath222 .
the gluon propagator in fig .
[ hardfig ] is @xmath401 } % \ ] ] setting @xmath402 , the denominator becomes @xmath403 then using eqs .
( [ limitqplus ] ) and ( [ limitqminus ] ) ( in either order ) , the denominator becomes @xmath404 that is , the denominator does not contain factors of the small variables , so that it has a finite limit as the small variables tend to zero
. we can write @xmath222 as a product @xmath405 the transverse components of the polarization vectors have the form @xmath406 for @xmath407 stands for any of the momenta @xmath190 , @xmath191 , @xmath17 or @xmath189 . for @xmath408 or @xmath191 we are ( temporarily ) defining @xmath409 .
the polarization vector for the detected gluon has @xmath410 but has a non - zero minus component @xmath411 the polarization vector for the final - state gluon , which we take to be in @xmath400 gauge according to eq .
( [ aminusgauge ] ) , has a non - zero plus component @xmath412 thus neither the gluon propagator nor any of the polarization vectors contains a factor of a small variable in the denominator .
we now consider the transverse tensor @xmath413 as a function of the variables @xmath414 @xmath415 is a rational function of these arguments and has dimension @xmath416 and boost dimension @xmath417 , where @xmath97 gives the scaling under boosts in the @xmath418 direction .
that is , for any four - vector @xmath407 , @xmath419 since @xmath415 has @xmath420 and @xmath416 , it can be written as a function of a reduced number of arguments , each of which has @xmath421 : @xmath422 we are interested in the limit in which the last five arguments are small , and we know that none of these arguments occur as factors in the denominators .
thus when eqs .
( [ limitqplus ] ) and ( [ limitqminus ] ) hold , @xmath415 approaches the limiting value @xmath423 we thus confirm the claim made in sec . [ decompose ] that we can neglect @xmath201 , @xmath208 , @xmath203 , and @xmath207 in @xmath222 .
furthermore , since the limiting function is covariant under rotations about the @xmath418 axis , it must be a linear combination of the tensors @xmath424 , @xmath425 and @xmath426 .
each of these tensors multiplies a rational function of @xmath427 .
that is , the coefficient functions are rational functions of @xmath194 .
this is , of course , just the structure given in eq .
( [ calmresult ] ) .
the argument given above does not establish that the limiting function is nonzero , but this is what we found by calculation .
we now return to the issue of whether the transverse components of the hard subgraph dominate over other components , as claimed in sec .
[ decompose ] .
we consider @xmath428 and choose values other than @xmath429 for @xmath384 or @xmath12 or both .
now @xmath430 or @xmath431 are not possible : @xmath432 multiplies the soft subgraph with corresponding indices , call it @xmath433 , which vanishes for @xmath430 or @xmath431 because of the gauge condition @xmath434 ( that is , @xmath435 ) .
thus we should consider @xmath436 or @xmath437 .
let us consider @xmath436 with @xmath438 as an example that illustrates the general argument .
thus we wish to investigate whether @xmath439 is dominated by @xmath440 in the limit specified by eqs .
( [ limitqplus ] ) and ( [ limitqminus ] ) .
we need an order of magnitude estimate for the ratio of @xmath441 to @xmath442 .
an analysis of lowest order graphs indicates that this is the same as the ratio of the corresponding components of the polarization vectors for a typical slow gluon , @xmath443 .
we write this as @xmath444 and compare @xmath445 to @xmath446 .
the analysis is simple .
@xmath447 has dimension @xmath416 and boost dimension @xmath448 .
thus it can be written as @xmath449 times a function @xmath450 that has dimension @xmath416 and boost dimension @xmath451 and , like @xmath452 , has no factors of the small variables in its denominator .
as our previous analysis shows , @xmath450 has a finite limit as the the small variables tend to zero .
( actually , @xmath450 vanishes in this limit because of its spin structure , but we will not need to use this fact . ) we note that the factor that multiplies @xmath450 @xmath453 vanishes in the limit specified by eq .
( [ limitqminus ] ) .
this establishes the claim .
in this appendix we compute the diffractive gluon distribution @xmath454 in a simple model . as we will see , this model does not exhibit the pomeron behavior @xmath455 with @xmath456 .
nevertheless , the model provides a check that @xmath457 can have a finite limit as @xmath458 in an exact numerical evaluation of the graphs to a given order of perturbation theory .
the model we will use is scalar - quark qcd , given by the lagrangian , @xmath459 this model has soft and collinear singularities just as in qcd .
as one further simplification , we model the diffractively scattered particle by the scalar @xmath460 field with a @xmath461 interaction to the quarks . then the perturbative @xmath462 interaction plays the role of the nonperturbative bethe - salpeter wave function of a real qcd meson . the probability for finding a @xmath463-pair inside the meson in this model falls of as @xmath464 in the ultraviolet , where @xmath465 is the transverse momentum of the quark .
this good behavior in the ultraviolet is due to the fact that g carries the dimension of mass .
thus the model allows for a simple treatment that approximately simulates properties of bound quarks inside a hadron . using this model , we compute the diffractive gluon distribution function within a meson , @xmath466 , as given in eq .
( [ fdifftog ] ) , to the lowest nontrivial order in @xmath467 , order @xmath468 .
one of the order @xmath468 diagrams for the function @xmath469 in eq .
( [ fdifftog ] ) is shown in fig .
[ figmod ] .
one gluon is detected , and a single gluon is exchanged between the two quark loops on opposite sides of the final state cut .
this gluon carries that portion of the momentum transfer from the meson that is lost to the detected gluon ( and in a physical process would be lost to the hard interaction ) .
the heavy bar ending with gluon lines represents the gluon density operator in eq .
( [ fdiff1 ] ) .
in addition to the graph shown , at each loop the gluons can also attach to the lower quark line . also in both cases there is a graph where the gluon lines are crossed .
in addition one gluon can attach on each of the quark lines .
finally there are the two - gluon two - quark contact interaction graphs . in total
this implies @xmath470 combinations .
after symmetry considerations , there are four types of amplitudes that must be explicitly evaluated .
the four amplitudes are shown in fig .
[ figblobs ] .
the final form for @xmath471 that we obtain is , @xmath472 where @xmath473 , @xmath474 and @xmath475 is a second rank tensor containing the sum of quark loops with all possible combinations of attachments by two gluons .
we have evaluated the loop integrals in eq .
( [ pertdist ] ) in terms of their explicit spence function expressions . for this
we have used the computer code of g. j. van oldenborgh and j. a. m.vermaseren @xcite .
their method is an independently formulated extension of the form algorithm of g. t hooft and m. veltman @xcite .
the authors of @xcite claim that their algorithms provide easier isolation of both asymptotic behavior and potential numerical instabilities . in one of the most sensitive regions , that of low @xmath7 , their algorithms have proven to be more accurate . to summarize the specifications of our calculation ,
it was for small @xmath7 , one external line was massless , and one had to integrate over a large region in the transverse space of the variable @xmath181 . due to gauge invariance , several constraints are placed on @xmath476 which we have numerically checked to hold .
@xmath475 has the form @xmath477 here the @xmath478 are tensors made from the available vectors @xmath479 .
the @xmath480 are scalar functions of the dot products of these momenta . in general
there would be ten terms in the sum , but gauge invariance cuts this in half . to find the possible @xmath478 ,
define @xmath481 notice that @xmath482 , _
etc_. now define @xmath483 these have the property that they are polynomials in the components of the momenta and that @xmath484 and @xmath485 .
thus @xmath476 is properly gauge invariant . in our calculation
, we transformed @xmath476 into a basis where the first five tensor forms are those given above and the other five are @xmath486 we have checked over a wide range of the parameters @xmath9 , @xmath7 and the transverse momentum @xmath181 that the coefficients of the tensors in eq .
( [ gauge2 ] ) vanish , leaving only the tensors in eq .
( [ gauge1 ] ) with nonvanishing coefficients . as a second check for gauge invariance of @xmath476
, we have numerically checked that @xmath487 and @xmath488 . in fig .
[ figcurves ] we show the diffractive gluon distribution function multiplied by @xmath489 , at @xmath490 , and for a selected choice of @xmath9 values .
the masses of the quarks and mesons were @xmath491 , while the couplings were @xmath492 and @xmath493 .
we see that the model does _ not _ exhibit the behavior @xmath494 for @xmath495 at fixed @xmath12 that would be characteristic of pomeron exchange .
one needs at least one rung of a gluon ladder to obtain this behavior .
the model exhibits the behavior @xmath496 as @xmath136 at fixed @xmath9 , as in eq .
( [ ginpomresult ] ) .
one can understand the @xmath12 and @xmath497 behavior seen in this numerical study from an analytic viewpoint . in sec .
[ tauglue ] , the detected and emitted gluons coupled do a slow gluon . in the present simple model , they couple directly to a fast quark
this means that the virtual quark line in fig .
[ figmod ] is far off shell when @xmath498 .
for example , it is off shell for @xmath499 for any @xmath12 as long as we take @xmath500 .
this gives quite a different behavior from that found in sec .
[ tauglue ] . in the model
, the loop serves as a built - in low momentum cut - off .
however , the meson can be replaced by a single fast quark line if we substitute an artificial low momentum cut - off @xmath501 then a simple power counting analysis gives , @xmath502\}^2 } % \ ] ] for @xmath500 and @xmath503 .
performing the integral gives @xmath504 this agrees with our numerical findings .
we find it reassuring that a fully consistent field theoretic calculation of the diffractive gluon distribution gives results that agree with expectations similar to our analytic arguments in sec .
[ tauglue ] .
zeus collaboration ( m. derrick , _ et al .
_ ) , phys . lett . * b315 * , 481 ( 1993 ) ; * b332 * , 228 ( 1994 ) ; * b338 * , 483 ( 1994 ) ; preprint desy 95 - 093 ( archive : [email protected] - 9505010 ) ; h1 collaboration , ( t. ahmed _ et al .
_ ) , nucl . phys .
* b429 * , 477 ( 1994 ) ; phys .
* b348 * , 681 ( 1995 ) .
g. ingelman and p. schlein , phys .
* 152b * , 256 ( 1985 ) .
e. l. berger , j. c. collins , d. e. soper , and g. sterman , nucl .
phys . * b286 * , 704 ( 1987 ) .
a. brandt , _ et .
* b297 * , 417 ( 1992 ) .
a. berera and d. e. soper , phys .
d * 50 * , 4328 ( 1994 ) .
g. veneziano and l. trentadue , phys .
* b323 * , 201 ( 1993 ) .
a. donnachie and p. v. landshoff , phys .
b * b191 * , 309 ( 1987 ) ; nucl . phys . * b303 * , 634 ( 1988 ) .
see , for example , e. gotsman , e. m. levin , u. maor , z. phys .
c * 57 * , 677 ( 1993 ) ; j. bartels and e. levin , nucl .
* b387 * , 617 ( 1992 ) ; j. bartels , nucl .
* a546 * , 107c ( 1992 ) . j. c. collins and d. e. soper , nucl .
* b194 * , 445 ( 1982 ) . g. curci , w. furmanski and r. petronzio , nucl .
* b175 * , 27 ( 1980 ) . v. n. gribov and l. n. lipatov , sov .
j. nucl .
phys . * 15 * , 438 ( 1972 ) ; l. n. lipatov , sov .
j. nucl .
phys . * 20 * , 93 ( 1975 ) ; yu .
l. dokshitzer , sov .
jept * 56 * , 641 ( 1977 ) ; g. altarelli and g. parisi , nucl .
* b26 * , 298 ( 1978 ) .
j. c. collins , d. e. soper , and g. sterman , nucl .
b261 * , 104 ( 1985 ) ; * b308 * , 833 ( 1988 ) . g. bodwin , phys . rev .
d * 31 * , 2616 ( 1985 ) ; * 34 * , 3932 ( 1986 ) . j. c. collins , l. frankfurt , and m. strikman , phys .
* b307 * , 161 ( 1993 ) .
a. donnachie and p. v. landshoff , nucl .
* b231 * , 189 ( 1984 ) ; * b244 * , 322 ( 1984 ) . j. c. collins , in _ physics simulations at high energy _ edited by v. barger , t. gottschalk , and f. halzen ( world scientific , singapore , 1987 ) .
v. s. fadin , e. a. kuraev , and l. n. lipatov , phys .
* b60 * , 50 ( 1975 ) ; i. balitsky and l. n. lipatov , sov .
j. nucl .
phys . * 15 * 438 ( 1978 ) .
h1 collaboration ( t. ahmed _ et al .
_ ) , nucl . phys .
* b439 * , 471 ( 1995 ) ; zeus collaboration ( m. derrick , _ et al .
_ ) , z. phys . *
c65 * , 379 ( 1995 ) .
a. capella , a. kaidalov , c. merino , d. pertermann and j. tran thanh van , preprint lpthe - orsay-95 - 33 , may 1995 , archive : [email protected] - 9506454 ; k. golec - biernat and j. kwiecinski , phys .
* b353 * , 329,(1995 ) . s. j. brodsky and g. farrar , phys .
lett . * 31 * , 1153 ( 1973 ) ; phys .
d * 11 * , 1309 ( 1975 ) ; v. a. matveev , r. m. murddyan and a. v. tavkheldize , lett .
nuovo cimento * 7 * , 719 ( 1973 ) ; f. close and d. sivers , phys .
* 39 * , 1116 ( 1977 ) ; d. sivers , ann .
nucl . part .
sci . * 32 * 149 ( 1982 ) . | diffractive parton distribution functions give the probability to find a parton in a hadron if the hadron is diffractively scattered .
we provide an operator definition of these functions and discuss their relation to diffractive deeply inelastic scattering and to photoproduction of jets at hera .
we perform a calculation in the style of `` constituent counting rules '' for the behavior of these functions when the detected parton carries almost all of the longitudinal momentum transferred from the scattered hadron . epsf.tex ( # 1 width # 2)=#2 |
nonnegative datasets are everywhere ; from term by document matrix induced from a document corpus @xcite , gene expression datasets @xcite , pixels in digital images @xcite , disease patterns @xcite , to spectral signatures from astronomical spectrometers @xcite among others .
even though diverse , they have one thing in common : all can be represented by using nonnegative matrices induced from the datasets .
this allows many well - established mathematical techniques to be applied in order to anayze the datasets .
there are many common tasks associated with these datasets , for example : grouping the similar data points ( _ clustering _ ) , finding patterns in the datasets , identifying important or interesting features , and finding sets of relevant data points to queries ( _ information retrieval _ ) . in this paper
, we will focus on two tasks : clustering and latent semantic indexing a technique that can be used for improving recall and precision of an information retrieval ( ir ) system .
clustering is the task of assigning data points into clusters such that similar points are in the same clusters and dissimilar points are in the different clusters .
there are many types of clustering , for example supervised / unsu - pervised , hierarchical / partitional , hard / soft , and one - way / many - way ( two - way clustering is known as co - clustering or bi - clustering ) among others . in this paper ,
clustering term refers to unsupervised , partitional , hard , and one - way clustering .
further , the number of cluster is given beforehand .
the nmf as a clustering method can be traced back to the work by lee & seung @xcite .
but , the first work that explicitly demonstrates it is the work by xu et al .
@xcite in which they show that the nmf outperforms the spectral methods in term of _ purity _ and _ mutual information _ measures for reuters and tdt2 datasets .
clustering aspect of the nmf , even though numerically well studied , is not theoretically well explained .
usually this aspect is explained by showing the equivalence between nmf objective to either k - means clustering objective @xcite or spectral clustering objective @xcite .
the problem with the first approach is there is no obvious way to incorporate the nonnegativity constraints into k - means clustering objective . and
the problem with the second approach is it discards the nonnegativity constraints , thus is equivalent to finding stationary points on unbounded region .
accordingly , the nmf which is a bound - constrained optimization turns into an unbounded optimization , so there is no guarantee the stationary point being utilized in proving the equivalence is located on the feasible region indicated by the constraints . in the first part of this paper
, we will provide a theoretical support for clustering aspect of the nmf by analyzing the objective at the stationary point using the karush - kuhn - tucker ( kkt ) conditions without setting the kkt multipliers to zeros .
thus , the stationary point under investigation is guaranteed to be located on the feasible region .
latent semantic indexing ( lsi ) is a method introduced by deerwester et al .
@xcite to improve recall and precision of an ir system using truncated singular value decomposition ( svd ) of the term - by - document matrix to reveal hidden relationship between documents by indexing terms that are present in the similar documents and weakening the influences of terms that are mutually present in the dissimilar documents .
the first capability can solve the _
synonymy_different words with similar meaning
problem , and the second capability can solve the _ polysemy_words with multiple unrelated meanings problem .
thus , lsi not only is able to retrieve relevant documents that do not contain terms in the query , but also can filter out irrelevant documents that contain terms in the query .
lsi aspect of the nmf is not well studied .
there are some works that discuss the relationship between the nmf and probabilistic lsi , e.g. , @xcite .
but the emphasize is in clustering capability of probabilistic lsi , not lsi aspect of the nmf .
motivated by the svd which is the standard method in clustering and lsi , in the second part of this paper , lsi aspect of the nmf will be studied , and the results will be compared to the results of the svd .
the nmf was popularized by the work of lee & seung @xcite in which they showed that this technique can be used to learn parts of faces and semantic features of text .
previously , it has been studied under the term positive matrix factorization @xcite .
mathematically , the nmf is a technique that decomposes a nonnegative data matrix into a pair of other nonnegative matrices : @xmath0 where @xmath1 $ ] denotes the data matrix , @xmath2 $ ] denotes the basis matrix , @xmath3 $ ] denotes the coefficient matrix , and @xmath4 denotes the number of factors which usually is chosen so that @xmath5 .
note that the definitions of @xmath6 , @xmath7 , and @xmath8 are chosen to simplify the interpretation of the nmf . to compute @xmath7 and @xmath8 , usually eq .
[ ch3:eq1 ] is rewritten into a minimization problem in frobenius norm .
@xmath9 in addition to the usual frobenius norm , family of bregman divergences which frobenius norm and kullback - leibler divergence are part of it can also be used as the distance measures . detailed discussion on bregman divergences can be found in , e.g. , ref . @xcite . in this work , we will consider only frobenius norm .
all nmf algorithms are formulated in the alternating fashion , fixing one matrix while solving the other ( the popular lee & seung algorithms @xcite and their derivatives , e.g. , @xcite also use the alternating strategy , but can not be represented by generic algorithm below ) .
this strategy is employed because the nmf is nonconvex with respect to @xmath7 and @xmath8 , but is convex with respect to @xmath7 or @xmath8 @xcite .
thus , the alternating strategy transforms nmf problem into a pair of convex subproblems . transforming a nonconvex problem into the corresponding convex subproblems
is a common practice in optimization researches because : ( 1 ) convex optimization is more tractable , ( 2 ) usually convex methods are more efficient , ( 3 ) any local optimum is necessarily a global optimum , and ( 4 ) the algorithms are easy to initialize @xcite .
initialization : @xmath10 .
@xmath11 algorithm [ ch3:alg1 ] solves nmf problem in the alternating fashion which will generate a solution sequence @xmath12 .
this algorithm is known as alternating nonnegativity - constrained least square ( anls ) algorithm , and usually is solved by decomposing each subproblem into the corresponding nonnegativity - constrained least square ( nnls ) problems , where there are many algorithms that guarantee the global - optimality of the nnls problems .
the following equations are the nnls versions of the anls in algorithm [ ch3:alg1 ] : @xmath13 where @xmath14 is the @xmath15-th row of @xmath16 . according to grippo & sciandrone @xcite , any limit point of @xmath17 , @xmath18generated by any anls algorithm that optimally solves the convex subproblem eq .
[ ch3:eq3 ] and eq .
[ ch3:eq4]is a stationary point . and
such anls based nmf algorithms exist , e.g. , @xcite , therefore there is guarantee that the stationary points are reachable . and as nnls is the building block for anls , any nnls algorithm that guarantees to find optimal solutions of eq .
[ ch3:eq5 ] and eq .
[ ch3:eq6 ] , e.g. , @xcite can also be employed to search for the stationary points . and as will be shown in section [ ch3:clusteringnmf ] , nmf objective ( eq . [ ch3:eq2 ] ) implicitly puts upper bounds on the feasible region ( the lower bounds are explicit : the nonnegativity constraints ) . thus the nmf is bound - constrained optimization problem , consequently @xmath17 , @xmath18 has at least one limit point @xcite .
this completes the conditions for any nmf algorithm that optimally solves subproblem eq .
[ ch3:eq3 ] and eq .
[ ch3:eq4 ] to have convergence guarantee .
this section is the first part of this paper in which a theoretical framework for supporting clustering aspect of the nmf will be provided .
the strict kkt optimality conditions will be utilized to derive the equivalence between nmf objective to graph clustering objective . unlike previous approaches where the kkt multipliers are set to zeros @xcite
, we will make no assumption about the kkt multipliers , thus the stationary point under investigation is guaranteed to be located on the feasible region in the nonnegative orthant .
we will also show that the feasible region is bounded , with the lower bounds are explicitly bounded by the nonnegativity constraints , and the upper bounds are implicitly bounded by the objective . as stated in section [ ch3:limitpoint ] ,
the boundedness of the feasible region is the necessary condition for guaranteeing the existence of limit point of @xmath19 . and
for interpretability reason , the data matrix @xmath6 will be considered as a feature - by - item data matrix unless stated differently .
the following proposition gives the theoretical support for clustering aspect of the nmf .
[ ch3:prop1 ] minimizing the following objective @xmath20 leads to the feature clustering indicator matrix @xmath7 and the item clustering indicator matrix @xmath8 .
@xmath21 since @xmath6 is constant , minimizing @xmath22 is equivalent to simultaneously optimizing : @xmath23 note that because @xmath24 , minimizing eq .
[ ch3:eq11 ] is equivalent to : @xmath25 the kkt function of objective in eq .
[ ch3:eq9 ] is : @xmath26 where @xmath27 and @xmath28 are the kkt multipliers . by applying the kkt optimality conditions to @xmath29
we get : @xmath30 with complementary slackness : @xmath31 where @xmath32 denotes component - wise multiplications .
eq . [ ch3:eq15 ] and eq . [ ch3:eq16 ]
lead to : @xmath33 substituting eq .
[ ch3:eq18 ] to eq .
[ ch3:eq10 ] leads to : @xmath34 which is equivalent to simultaneously optimizing : @xmath35 similarly , substituting eq .
[ ch3:eq17 ] to eq .
[ ch3:eq10 ] leads to : @xmath36 which is equivalent to simultaneously optimizing : @xmath37 as shown , eq .
[ ch3:eq22 ] and eq .
[ ch3:eq26 ] recover eq .
[ ch3:eq12 ] and eq .
[ ch3:eq13 ] respectively , so there is no need to substituting eq . [ ch3:eq17 ] and eq . [ ch3:eq18 ] into eq .
[ ch3:eq11 ] .
now we concentrate on the basis matrix @xmath7 first . eq .
[ ch3:eq20 ] [ ch3:eq22 ] give alternative objectives to the original nmf objective that contain only @xmath7 . note that if we consider @xmath6 to be an affinity matrix induced from bipartite graph @xmath38 ( which is a reasonable thought since any feature - by - item matrix can be modeled by a bipartite graph ) , then @xmath39 is the feature graph where edge weights describe the similarity between corresponding vertex pairs .
so , eq . [ ch3:eq20 ] looks like _ ratio association _ applied to @xmath39 . but without orthogonality constraint @xmath40 ( which is the part of _ ratio association _ objective ) , one can optimize eq . [ ch3:eq20 ] by setting @xmath7 to an infinity matrix
. however , this violates eq .
[ ch3:eq22 ] which favours small @xmath7 .
similarly , one can optimize eq .
[ ch3:eq22 ] by setting @xmath7 to a zero matrix . but
again , this violates eq .
[ ch3:eq20 ] .
thus , eq . [ ch3:eq20 ] and eq . [ ch3:eq22 ] create implicit lower and upper bound constraints on @xmath7 : @xmath41 . for convenience , eq .
[ ch3:eq22 ] can be restated as : @xmath42 by using the fact @xmath43 , eq .
[ ch3:eq27 ] can be rewritten into : @xmath44 therefore , eq . [ ch3:eq20 ] [ ch3:eq22 ] can be restated as : @xmath45 even though @xmath7 is now bounded , since there is no column - orthogonality constraint , maximizing eq .
[ ch3:eq29 ] can be easily done by setting each entry of @xmath7 to the corresponding largest possible value ( in graph term this means to only create one partition on @xmath46 ) . but
this scenario results in a large value of eq .
[ ch3:eq31 ] , which violates the objective .
similarly , minimizing eq .
[ ch3:eq31 ] to the smallest possible value violates eq .
[ ch3:eq29 ] . since minimizing @xmath47 implies minimizing @xmath48 , but not vice versa , simultaneously optimizing eq .
[ ch3:eq29 ] and eq .
[ ch3:eq31 ] can be done by setting @xmath48 as small as possible and balancing @xmath47 with eq .
[ ch3:eq29 ] .
this scenario is the relaxed _ ratio association _ applied to @xmath46 , and as long as vertices in @xmath46 are clustered , it leads to the grouping of related features .
the remaining problem is eq .
[ ch3:eq30 ] .
since we know nothing about @xmath49 , the best bet will be making each entry of @xmath50 as large as possible .
this can be done by setting @xmath7 to the largest possible values , but this scenario violates eq .
[ ch3:eq31 ] .
so , the most reasonable scenario will be making the entries near diagonal region of @xmath50 as large as possible .
this can be achieved by using @xmath7 from previous discussion . as @xmath7 is the feature clustering indicator matrix , multiplying @xmath51 with @xmath7 will result in a matrix that has larger entries near diagonal region ,
therefore it can be expected that eq .
[ ch3:eq30 ] will have good optimality .
thus simultaneously optimizing eq .
[ ch3:eq29 ] [ ch3:eq31 ] leads to the feature clustering indicator matrix @xmath7 . by applying the similar approach to the coefficient matrix @xmath8 , optimizing eq .
[ ch3:eq24 ] [ ch3:eq26 ] is equivalent to optimizing : @xmath52 where @xmath53 denotes @xmath15-th row of @xmath8 . by following the previous discussion on @xmath7
, it can be shown that as long as vertices in @xmath54 are clustered , simultaneously optimizing eq .
[ ch3:eq32 ] [ ch3:eq34 ] leads to the item clustering indicator matrix @xmath8 . as shown in the proof of proposition [ ch3:prop1 ] , optimizing nmf objective is equivalent to applying the relaxed _ ratio association _ to the item graph @xmath54 and the feature graph @xmath39 simultaneously . and because in the nmf , clustering membership of each point is directly determined by finding the largest projection on the axis of the decomposition rank subspace ( @xmath4 subspace ) @xcite , the nmf can only offer good results if the data points are linearly separable .
this is not the case with the spectral clustering , where the memberships are indirectly determined by applying k - means clustering on the resulting factors .
this additional step can sometimes find correct assignments even though the data points are not linearly separable . and
unfortunately , since the factors produced by the nmf are nonnegative and directly point to the cluster s centers @xcite , applying k - means clustering on the factors wo nt change the clustering assignments .
the following examples show the limitation of the nmf in clustering linearly inseparable data points . and for comparison , the spectral clustering is used .
for the spectral clustering , we use ng et al .
algorithm ( njw ) @xcite , and for the nmf , we use lee & seung algorithm ( nmfls ) @xcite , and kim & park algorithm ( nmfjk ) @xcite .
njw and nmfls are the standard algorithm for the spectral clustering and the nmf respectively , and nmfjk is the nmf algorithm that has convergence guarantee .
algorithm [ ch2:alg1 ] describes njw algorithm , and algorithm [ ch3:alg2 ] describes clustering using the nmf .
note that we wrote codes for njw and nmfls by ourselves , and use codes from the authors website for nmfjk . to get the same treatment as in njw
, we use the same kernel strategy for nmfls and nmfjk .
the adjustable parameter @xmath55 is learned directly from the datasets , and the results are displayed in figure [ ch2:fig1 ] , [ ch3:fig1 ] , and [ ch3:fig2 ] . 1 .
input : rectangular data matrix @xmath56 with @xmath57 data points , # cluster @xmath4 , and gaussian kernel parameter @xmath55 .
2 . construct symmetric affinity matrix @xmath58 from @xmath6 by using gaussian kernel .
3 . normalize @xmath59 by @xmath60 where @xmath61 is a diagonal matrix with @xmath62 .
4 . compute the @xmath4 largest eigenvectors of @xmath59 , and form @xmath63 $ ] , where @xmath64 is the @xmath65-th largest eigenvector of @xmath59 .
normalize every row of @xmath66 , i.e. , @xmath67 .
apply k - means clustering on the row of @xmath66 to obtain the clustering indicator matrix @xmath68 . 1 .
input : rectangular data matrix @xmath56 with @xmath57 data points , # cluster @xmath4 , and gaussian kernel parameter @xmath55 .
2 . construct symmetric affinity matrix @xmath58 from @xmath6 by using gaussian kernel .
3 . compute @xmath7 and @xmath8 by using nmf algorithm ( nmfls or nmfjk ) so that @xmath69 .
assume @xmath8 is used , then clustering assignment of data point @xmath70 , @xmath71 , can be computed by @xmath72 . as shown in figure [ ch2:fig1 ] , [ ch3:fig1 ] , and [ ch3:fig2 ] , while the spectral clustering can correctly find the clustering assignments for all datasets , the nmfs can only compete with the spectral clustering for the last dataset which is rather linearly separable .
these results are in accord with the proof of proposition [ ch3:prop1 ] ( that states as long as vertices on the feature @xmath73item@xmath74 graph are clustered , optimizing the nmf objective leads to the feature @xmath73item@xmath74 clustering indicator matrix ) .
thus , it seems that as a clustering method , the nmf is more similar to k - means clustering or support vector machine ( svm ) which also can only cluster linearly separable datasets , than to the spectral methods , even though both clustering using the nmf and the spectral methods are based on matrix decomposition techniques .
accordingly , clustering performances of the nmf can probably be improved by using appropriate kernel methods as in k - means clustering and svm .
the experiments are conducted to evaluate the performances of the nmf as a clustering method .
all algorithms are developed in gnu octave under linux platform using a notebook with 1.86 ghz intel processor and 2 gb ram .
reuters-21578 document corpus , the standard dataset for testing learning algorithms and other text - based processing methods , is used for this purpose .
this dataset contains 21578 documents ( divided into 22 files with each file contains 1000 documents and the last file contains 578 documents ) with 135 topics created manually with each document is assigned to one or more topics based on its content .
the dataset is available in sgml and xml format , we use the xml version .
we use all but the 18@xmath75 file because this file is invalid both in its sgml and xml version .
we use only documents that belong to exclusively one class ( we use `` classes '' for refeering to the original grouping , and `` clusters '' for referring to groups resulted from the clustering algorithms ) .
further , we remove the common english stop words , stem the remaining words using porter stemmer @xcite , and then remove words that belong to only one document . and also , we normalize the term - by document matrix @xmath6 by : @xmath76 where @xmath77 as suggested by xu et al .
we form test datasets by combining top 2 , 4 , 6 , 8 , 10 , and 12 classes from the corpus .
table [ ch2:table3 ] summarizes the statistics of these test datasets , where # doc , # word , % nnz , max , and min refer to the number of document , the number of word , percentage of nonzero entry , maximum cluster size , and minimum cluster size respectively . and table [ ch2:table4 ] gives the sizes ( # doc ) of these top 12 classes .
.statistics of the test datasets . [ cols="<,>,>,>,>,>",options="header " , ] [ ch3:table13 ]
we have presented a theoretical framework for supporting clustering aspect of the nmf without setting the kkt multipliers to zeros . thus the stationary point used in proving this aspect
is guaranteed to be on the nonnegative orthant which is the feasible region of the nmf .
our theoretical work implies a limitation of the nmf as a clustering method in which it can not be used in clustering linearly inseparable datasets .
so , the nmf as a clustering method is more resembling k - means clustering or svm than the spectral clustering , even though both the nmf and the spectral methods utilize matrix decomposition techniques . as the clustering capabilities of k - means and svm usually can be improved by using the kernel methods , probably the same approach can also be employed in the nmf .
we will address this issue in our future researches . clustering capability of nmfjk is comparable to the svd in reuters datasets with nmfjk tends to be better for small # cluster and the svd for big # cluster .
but unfortunately , nmfls which is the standard nmf algorithm can not outperform the svd .
these results imply clustering aspect of the nmf is algorithm - dependent , a fact that seems to be overlooked in the nmf researches .
lsi aspect of the nmf seems to be comparable to the svd in its power for solving synonymy and polysemy problems for datasets with clear semantic structures that allowed these problems to be revealed . in real datasets , however , the nmf generally can not outperform the svd .
but an interesting fact comes into sight ; in some cases , the nmf can outperform the svd , even though when the computations are repeated and averaged over the number of trials , these advantages vanish . because the nmf can offer different results depending on the algorithms , the initializations , the objectives , and the problems , improving lsi capability of the nmf is possible .
we will address this problem in our future researches .
h. kim and h. park , `` sparse non - negative matrix factorizations via alternating non - negativity constrained least squares for microarray data analysis , '' bioinformatics , vol .
23(12 ) , pp . 1495 - 502 , 2007 . c. ding , t. li , and w. peng , `` on the equivalence between non - negative matrix factorization and probabilistic latent semantic indexing , '' computational statistics & data analysis , vol .
52(8 ) pp . 3913 - 27 , 2008 . h. kim and h. park , `` nonnegative matrix factorization based on alternating nonnegativity constrained least squares and active set method , '' siam .
j. matrix anal . & appl .
30(2 ) , pp . 713 - 30 , 2008 .
r. albright , j. cox , d. duling , a. langville , and c. meyer , `` algorithms , initializations , and convergence for the nonnegative matrix factorization , '' ncsu technical report math 81706 , north carolina state university , 2006 . | this paper provides a theoretical support for clustering aspect of the nonnegative matrix factorization ( nmf ) . by utilizing the karush - kuhn - tucker optimality conditions ,
we show that nmf objective is equivalent to graph clustering objective , so clustering aspect of the nmf has a solid justification .
different from previous approaches which usually discard the nonnegativity constraints , our approach guarantees the stationary point being used in deriving the equivalence is located on the feasible region in the nonnegative orthant .
additionally , since clustering capability of a matrix decomposition technique can sometimes imply its latent semantic indexing ( lsi ) aspect , we will also evaluate lsi aspect of the nmf by showing its capability in solving the synonymy and polysemy problems in synthetic datasets . and more extensive evaluation will be conducted by comparing lsi performances of the nmf and the singular value decomposition ( svd)the standard lsi method using some standard datasets .
bound - constrained optimization , clustering method , nonnegative matrix factorization , karush - kuhn - tucker optimality conditions , latent semantic indexing . 15a23 , 68r10 . |
over the next few years a number of ground based , balloon and satellite experiments will measure cmb sky with an unprecedented accuracy and detail .
the promise of a one percent precision on the measured power spectrum of cmb anisotropies requires a similar accuracy in the theoretical predictions , if we are to exploit all the information present in the data .
the rewards will be rich : among other things this will allow an accurate determination of a number of cosmological parameters and testing of current structure formation theories @xcite . in principle
such a program is possible , since the anisotropies were produced when the universe was still in the linear regime , which makes the calculations of model predictions very accurate . in practice
there are a number of important effects that need to be included if this goal is to be realized .
one of the most important among these is the gravitational lensing effect .
as photons propagate through the universe from their last scattering to our detectors they are randomly deflected by the gravitational force exerted upon them by the inhomogeneous mass distribution .
previous work has shown that gravitational lensing has an effect on the temperature anisotropy power spectrum which is not insignificant @xcite .
the random deflections smear out the sharp features in the correlation function or power spectrum , leading to a suppression of acoustic oscillations .
gravitational lensing can also enhance power on the damping tail , causing it to decay less rapidly than predicted on very small angular scales @xcite .
gravitational lensing effect on the temperature anisotropies has been discussed several times in the literature and the formalism to calculate it using the evolution of density power spectrum both in linear and nonlinear regime has been presented in @xcite . in this paper
we extend this calculation to the two linear polarization power spectra and to the cross - correlation spectrum between temperature and polarization . because acoustic oscillations are narrower for polarization spectra than for temperature , one expects gravitational lensing effect to be more significant in the former and indeed our results confirm this .
in addition , a qualitatively new effect is the mixing between @xmath2 and @xmath1 types of polarization , which changes the pattern of polarization .
the outline of the paper is the following . in [ formalism ]
we develop the formalism : this section contains all the main analytic expressions needed for a numerical implementation of the effect .
these have been numerically implemented in the new version of cmbfast package ( version 2.4 ) and require only a marginal increase in the cpu time for their evaluations . in [ estimate ]
we compute the effect for a typical cosmological model and address the question of direct observability of the effect .
we present the conclusions in [ conclusions ] .
the large scale density fluctuations in the universe induce random deflections in the direction of the cmb photons as they propagate from the last scattering surface to us .
this alters the power spectrum of both the temperature and polarization anisotropies .
the quantity is responsible for the deflections is the projected surface density . since the structures are not very correlated on large scales the gravitational lensing effect is only relevant at the small angular scales in the cmb . we may hence use the small scale limit formalism @xcite , which simplifies the calculations . the observed cmb temperature in the direction @xmath3 is @xmath4 and equals the unobservable temperature at the last scattering surface @xmath5 , where @xmath6 is the angular excursion of the photon as it propagates from the last scattering surface until the present . in terms of fourier components we have @xmath7
the same relation applies to the two stokes parameters @xmath8 and @xmath9 that describe linear polarization @xcite . in the small scale limit
we take the direction of observation to be near @xmath10 and we orient the local coordinate system orthogonal to @xmath10 to define @xmath8 and @xmath9 : @xmath8 is the difference between the photon intensities along @xmath11 and @xmath12 , while @xmath9 is the difference between photon intensities along the two diagonals . in terms of fourier components these can be expressed with opposite parity fourier components @xmath13 and @xmath14 @xcite @xmath15 \nonumber \\ u({\bi
\theta})&=&{\tilde}u({\bi \theta}+\delta { \bi \theta } ) \nonumber \\ & = & ( 2\pi)^{-2}\int d^2{\bi l}\ e^{i{\bi l}\cdot ( { \bi \theta}+\delta { \bi \theta})}\ [ e({\bi l } ) \sin(2\phi_{\bi l } ) + b({\bi l } ) \cos(2\phi_{\bi l } ) ] .
\label{expansion } \end{aligned}\ ] ] the fourier components satisfy @xmath16 with @xmath17 and the average is over different realizations of the cmb field .
the correlation function of the temperature between two points in the sky , @xmath18 and @xmath19 , only depends on their angular separation @xmath20 . with our choice of coordinate system
the correlations between the polarization variables are also a function of azimuthal angle @xmath21 @xcite .
this is not the natural coordinate system in which to define these correlation functions .
instead one should align the local coordinate system with the great circle that connects the two points @xcite . with this choice parity conservation
requires @xmath22 and @xmath8 to be uncorrelated with @xmath9 and the correlation functions only depend on separation @xmath20 . to obtain this set of correlation functions in our coordinate system
we can calculate the correlation between the variables at the origin and another point separated by an angle @xmath20 along the @xmath23 axis .
the correlation functions are calculated from equations ( [ expansion ] ) and ( [ variance ] ) , @xmath24\nonumber \\ c_u({\theta})&=&\int { d^2{\bi l}\over ( 2\pi)^2}\ e^{il\theta \cos \phi_{l } } \langle e^{i{\bi l}\cdot ( \delta{\bi \theta^a } -\delta{\bi \theta}^{b})}\ \rangle\ \ [ c_{{\tilde}el } \sin^2(2\phi_{\bi l } ) + c_{{\tilde}bl } \cos^2(2\phi_{\bi l } ) ] \nonumber \\
c_c({\theta})&=&\int { d^2{\bi l}\over ( 2\pi)^2}\ e^{il\theta \cos \phi_{l } } \langle e^{i{\bi l}\cdot ( \delta{\bi \theta^a } -\delta{\bi \theta}^{b})}\ \rangle\ \ c_{{\tilde}cl } \cos(2\phi_{\bi l } ) .
\label{corrlens}\end{aligned}\ ] ] the remaining average in equation ( [ corrlens ] ) is over the lensing fluctuations .
only the cross - correlation between @xmath8 and @xmath22 is different from zero and we denote it @xmath25 . even in the presence of lensing @xmath9 does not cross correlate with either @xmath22 or @xmath8 because these quantities have opposite parities . the correlation function of the excursion angle can be used to calculate the expectation value in equation ( [ corrlens ] ) @xcite , @xmath26\ } \nonumber \\ & \approx & 1-{l^2 \over 2 } [ \sigma_0 ^ 2(\theta ) + \cos(2
\phi_{\bi l } ) \sigma_2 ^ 2(\theta ) ] .
\label{avgexp}\end{aligned}\ ] ] the exponential above has been expanded out assuming @xmath27 and @xmath28 are small and we numerically verified this to be an excellent approximation ( see also @xcite ) .
the two functions characterizing the rms dispersion of the photons are @xcite @xmath29 \nonumber \\ \sigma_2 ^ 2(\theta)&\equiv & 16 \pi^2 \int_0^\infty k^3 dk \int_0^{\chi_{rec } } p_{\phi}(k,\tau=\tau_0-\chi)\ w^2(\chi,\chi_{rec})\ j_2(k\theta\sin_k \chi ) .
\label{sigmalens}\end{aligned}\ ] ] we denote with @xmath30 the comoving radial distance , @xmath31 is the conformal time and @xmath32 corresponds to its value today .
the comoving angular diameter distance is @xmath33 in a closed universe ( @xmath34 ) , @xmath35 in a flat universe ( @xmath36 ) and @xmath37 in an open universe ( @xmath38 ) .
the curvature @xmath39 can be expressed using the present density curvature parameter @xmath40 and the present hubble parameter @xmath41 as @xmath42 .
the power spectrum of the gravitational potential at time @xmath31 is @xmath43 and @xmath44 , where @xmath45 is the radial distance to the last scattering surface at recombination .
we include the non linear evolution of the power spectrum using the fitting formulae of peacocks and dodds @xcite .
these expressions are thus valid in a general robertson - walker metric , both in linear and nonlinear regime .
the nonlinear effects are however not very important except on small angular scales .
we use equations ( [ avgexp ] ) and ( [ sigmalens ] ) together with ( [ corrlens ] ) to obtain the final expression for the correlation functions , @xmath46 + { l^2
\over 2}\sigma_2 ^ 2(\theta ) j_2(l\theta)\}\ \nonumber
\\ c_{q}(\theta)+c_{u}(\theta)&=&\int { l dl\over 2\pi } \
( c_{{\tilde}el}+c_{{\tilde}bl})\ \ { j_0(l\theta)[1-{l^2 \over 2}\sigma_0 ^ 2(\theta ) ] + { l^2 \over 2 } \sigma_2 ^ 2(\theta ) j_2(l\theta)\ } \nonumber \\
c_{q}(\theta)-c_{u}(\theta)&=&\int { l dl\over 2\pi } \ ( c_{{\tilde}el}-c_{{\tilde}bl})\ \ { j_4(l\theta)[1-{l^2 \over 2}\sigma_0 ^ 2(\theta ) ] + { l^2 \over 4 } \sigma_2 ^ 2(\theta ) [ j_2(l\theta)+j_6(l\theta)]\ } \nonumber
\\ c_c(\theta)&=&\int { l dl\over 2\pi } \ c_{{\tilde}cl}\ \ { j_2(l\theta)[1-{l^2 \over 2}\sigma_0 ^ 2(\theta ) ] + { l^2 \over 4 } \sigma_2 ^ 2(\theta ) [ j_0(l\theta)+j_4(l\theta)]\ } .
\label{fullcorr}\end{aligned}\ ] ] the power spectrum in fourier space has become the most widely used way to characterize the cmb anisotropies .
we can compute the different power spectra from the correlation functions , @xmath47\ j_0(l\theta ) + [ c_q(\theta)-c_u(\theta)]\ j_4(l\theta ) \ } \nonumber \\
c_{bl}&=&2\pi \int_0^\pi \theta d\theta \ \ { [ c_q(\theta)+c_u(\theta)]\ j_0(l\theta ) - [ c_q(\theta)-c_u(\theta)]\ j_4(l\theta ) \ } \nonumber \\
c_{cl}&=&2\pi \int_0^\pi \theta d\theta \
c_c(\theta)\ j_2(l\theta ) . \label{defcl}\end{aligned}\ ] ] equations ( [ fullcorr ] ) and ( [ defcl ] )
give the mapping between the observed cmb power spectra and the primordial one .
explicitly , @xmath48\
c_{{\tilde}el^{\prime } } + { 1 \over 2}[{\cal w}_{1 l}^{l^\prime}-{\cal w}_{2 l}^{l^\prime}]\ c_{{\tilde}bl^{\prime } } \nonumber \\
c_{bl}&=&c_{{\tilde}bl}+{1 \over 2}[{\cal w}_{1 l}^{l^\prime}-{\cal w}_{2 l}^{l^\prime}]\ c_{{\tilde}el^{\prime } } + { 1 \over 2}[{\cal w}_{1 l}^{l^\prime}+{\cal w}_{2 l}^{l^\prime}]\ c_{{\tilde}bl^{\prime } } \nonumber \\
c_{cl}&=&c_{{\tilde}cl}+{\cal w}_{3 l}^{l^\prime}\ c_{{\tilde}cl^{\prime } } , \label{cll}\end{aligned}\ ] ] the sum over @xmath49 is implicit .
the window functions are defined to be @xmath50 -\sigma_0 ^ 2(\theta)j_4(l^{\prime}\theta)\ } \nonumber \\ { \cal w}_{3 l}^{l^\prime}&=&{{l^\prime}^3 \over 2}\int_0^{\pi } \theta d\theta \
j_2(l\theta)\ \{{1 \over 2}\sigma_2 ^ 2(\theta)[j_0(l^{\prime}\theta ) + j_4(l^{\prime}\theta)]-\sigma_0 ^ 2(\theta)j_2(l^{\prime}\theta)\}. \label{clmapping}\end{aligned}\ ] ] equations ( [ cll ] ) and ( [ clmapping ] ) are the main result of the paper .
the results for polarization are new and represent a generalization of previous results for the temperature @xcite .
the important qualitatively new feature is that lensing mixes @xmath2 and @xmath1 polarization modes . on small scales where the lensing effect is important
all cosmological models proposed so far predict only @xmath2 type polarization and @xmath51 .
lensing will however generate @xmath1 type polarization in the observed field , @xmath52\ c_{{\tilde}el^{\prime}}\neq 0 $ ] . in the next section we will calculate the lensed power spectra in a typical cosmological model to address the significance of the effect . before addressing this issue
let us explore a simplified model to understand why the two polarization types are mixed through lensing .
so far we have introduced @xmath2 and @xmath1 type polarizations in fourier space .
we can also define real space quantities @xmath53 these two quantities describe completely the polarization field and it proves easier to understand the effect of lensing in terms of these .
they can also be computed directly from @xmath8 and @xmath9 in real space @xcite , @xmath54 we have defined @xmath55 and @xmath56 , the stokes parameters in the polar coordinate system centered at @xmath57 . if @xmath58 then @xmath59 and @xmath60 .
the window is @xmath61 , @xmath62 .
we consider as a toy model an unlensed polarization field which is a radial pattern around the origin in a ring of size @xmath63 .
we have @xmath64 and @xmath65 . in the absence of lensing
we would observe at the origin @xmath66 and @xmath67 .
this follows from @xmath68 ( because polarization is radial ) and equation ( [ ebreal ] ) .
we now consider what happens if we shift the position of the photons by a random angle of size @xmath69 .
each segment of the ring will be mapped to a different position with angle @xmath21 where @xmath70 .
each segment acquires a random component of non radial polarization and since we are assuming the segment shifts are uncorrelated the integral of @xmath56 over @xmath21 does not vanish .
it has mean zero and variance @xmath71 .
the measured power in the @xmath1 mode is @xmath72 .
this example shows that the pattern of polarization vectors on the sky determines the amount of @xmath2 and @xmath1 type polarization .
lensing distorts this pattern by shifting the positions of the photons in the plane of the sky relative to the last scattering surface .
it can thus generate @xmath1 pattern out of initially pure @xmath2 polarization . in the next section
we discuss the amplitude of this effect .
the effect of gravitational lensing on the temperature power spectrum has been studied in detail in previous work @xcite .
we will explore here the effects on polarization and asses the detectability of the @xmath1 signal . in these examples
we will use the cosmic concordance model @xcite with @xmath73 km s@xmath74mpc@xmath74 , @xmath75 , @xmath76 , @xmath77 and @xmath78 .
the model is cobe normalized and satisfies the constraints on 8 h@xmath74mpc scale from the abundance of clusters .
figure [ fig1 ] shows the functions that characterize the rms deflection of the photons , @xmath79 and @xmath80 . on small scales the rms relative deflection between two photons approaches 15% , justifying the weak lensing assumption over most of the sky ( for caustic formation one requires @xmath81 , which will occur only on rare occasions ) .
we also show @xmath82 $ ] for @xmath83 .
the two window functions @xmath84 and @xmath85 are very similar , making their difference much smaller than their sum .
one thus expects the generated @xmath1 type polarization from a pure @xmath2 type ( and viceversa ) to be very small .
in fact @xmath86 is also very similar , the three windows only differ at the 1% level .
the windows are oscillatory as a function of @xmath49 and their main contribution is concentrated around @xmath87 .
figure [ fig2 ] ( upper panel ) shows the lensed and unlensed power spectra . on intermediate scales
lensing has the general effect of smearing the peaks in the spectrum by redistributing power .
polarization receives contribution only from velocity gradients at the last scattering surface and the acoustic peaks are very sharp .
in contrast , temperature receives contributions both from velocity and density of photon - baryon plasma and the two are out of phase with each other , leading to a partial cancellation of acoustic peaks .
one thus expects the effect of lensing to be larger for polarization and indeed we find that it is approximately a factor of two larger for @xmath2 polarization ( figure [ fig2 ] , bottom left panel ) . in the damping tail
the power is enhanced over the unlensed case as can be seen in the general trend in the relative difference between lensed and unlensed spectra ( figure [ fig2 ] ) . to asses the detectability of the induced @xmath1 polarization we will focus on the plack mission .
we assume that measurements of the temperature , @xmath2 type polarization and @xmath88 cross correlation have allowed the determination of the cosmological parameters accurately enough so that we know the approximate shape of the @xmath1 power spectra induced by lensing .
we will attempt to determine just one parameter , the overall normalization of this @xmath1 signal , which expresses the overall effect of the matter fluctuations along the line of sight . if the signal to noise were large enough we could also attempt to determine the power spectrum of the lensing effect , by exploring the signal in @xmath1 as a function of @xmath87 . the relative error on the overall amplitude of the induced @xmath1 component @xmath89 is @xmath90 @xmath91 is the fraction of observed sky which we take to be @xmath92 .
we assume the noise and beam width to be @xmath93 and @xmath94 with @xmath95 .
the effective number of modes contributing information is @xmath96 , so that @xmath97 . for the experimental specifications above @xmath98 and @xmath99 .
this means that there are only 8 independent modes of @xmath1 type polarization that can be observed with planck , each a gaussian random variable with 0 mean and so the signal will be at the limit of detectability by planck .
the @xmath1 signal peaks at @xmath100 and a ground based experiment observing a small patch of the sky for sufficiently long time to reduce the noise per pixel would be more effective to detect this signal .
gravitational lensing induced @xmath1 is also not a significant contaminant of the @xmath1 polarization expected from tensor modes in inflationary models @xcite , the latter being dominant on large angular scales ( figure [ fig2 ] , bottom right panel , where we assumed tensors and scalars are of equal amplitude in the temperature on large scales ) .
the change in the cmb spectra induced by random deflections of the photons by the large scale structure of the universe has to be included in the calculations of the anisotropies when comparing theory and observations .
we have presented a formalism to calculate the effect of gravitational lensing on all four cmb power spectra .
the final expressions are given in a compact and numerically efficient form that is adequate for numerical implementation .
these expressions have been implemented in the cmbfast package , freely available from the authors .
the computational time does not significantly increase with this feature and since the effect can be significant we feel the lensing effect should be included in any calculation where high precision accuracy is important , such as in the design and analysis of the cmb experiments .
the method is self - consistent in the sense that for any cosmological model we use the actual power spectrum computed in the code to compute the lensing effect .
the power spectrum is normalized to cobe using the cmb spectrum computed from the same code output .
this approach gives the correct amplitude of gravitational lensing effect for the particular model in question .
it should be pointed out that if the model is inconsistent with the small scale constraints such as @xmath101 normalization then the amplitude of the lensing effect will also be incorrect .
for example , cobe normalized standard cdm has a much larger gravitational lensing effect than what we find using the cosmic concordance model , but this is only a reflection of standard cdm model having too much small scale power to be compatible with small scale constraints . for models that are correctly normalized on small scales the relative change of the polarization power spectra can reach 10% at @xmath102 and even more at higher @xmath87 .
this is larger than in the temperature spectrum and is caused by the sharper acoustic peaks in the polarization spectra . in the damping region the lensed spectra show an enhancement above the unlensed spectra just like in the temperature case , although the sensitivity to this effect is too small for satellite missions to detect it in polarization @xcite .
gravitational lensing also mixes @xmath2 and @xmath1 type polarization by deforming the polarization pattern on the sky relative to that at the last scattering surface
. this will generate @xmath1 type polarization out of the @xmath2 type polarization even if there was no @xmath1 present at the last scattering surface .
this induced @xmath1 mode is rather small in typical models and will be only marginally detectable by the planck surveyor .
it peaks at fairly small angular scales around @xmath103 and so does not affect the measurement of gravity waves from @xmath1 polarization on larger scales . a ground based experiment observing a small patch of the sky would be more suitable to observe this effect and would allow one to determine the power spectrum of matter fluctuations .
this would allow a detection of a combined gravitational effect from structures spanning a much larger range in redshift space than currently reachable by other methods .
mz was supported by nasa grant nag5 - 2816 .
jungman g. , kamionkowski m. , kosowsky a. , and spergel d. n. phys .
lett . , * 76 * , 1007 ( 1996 ) ; _ ibid _ phys .
d * 54 * , 1332 ( 1996 ) ; bond j. r. , efstathiou g. & tegmark m. , mnras , * 291 * , 33 ( 1997 ) ; zaldarriaga m. , spergel d. n. and seljak u. astrophys .
j. * 488 * , 1 ( 1997 ) blanchard , a. and schneider , j. , astron .
& astrophys .
* 184 * , 1 ( 1987 ) ; cole , s. and efstathiou , g. , mnras * 239 * , 195 ( 1989 ) ; cay ' on , l. , mart ' inez
- gonz ' alez , e. , & sanz , j. l. astrophys . j. * 403 * , 471 ( 1993 ) ; _ ibid _ astrophys .
j. * 489 * , 21 ( 1997 ) | we investigate the effect of gravitational lensing by matter distribution in the universe on the cosmic microwave background ( cmb ) polarization power spectra and temperature - polarization cross - correlation spectrum . as in the case of temperature spectrum gravitational lensing
leads to smoothing of narrow features and enhancement of power on the damping tail of the power spectrum . because acoustic peaks in polarization spectra are narrower than in the temperature spectrum the smoothing effect is significantly larger and can reach up to 10% for @xmath0 and even more above that .
a qualitatively new feature is the generation of @xmath1 type polarization even when only @xmath2 is intrinsically present , such as in the case of pure scalar perturbations .
this may be directly observed with planck and other future small scale polarization experiments .
the gravitational lensing effect is incorporated in the new version ( 2.4 ) of cmbfast code .
epsf # 1 |
since the early investigation of the antiferromagnetic structure of lacro@xmath0 ( @xmath6 320 k ) by koehler and wollan@xcite rare earth orthochromites with the chemical composition rcro@xmath0 , ( r = rare earth element ) have experienced particular attention .
interest was especially triggered by their complex magnetic ordering often with spin canting resulting from the interaction of the magnetic moments of the cr@xmath5 and the r@xmath5 cations which can be tuned by external parameters like temperature , magnetic field or pressure.@xcite as such the rare earth orthochromites were perceived as systems with great potential as magnetoelectric multiferroics ( me / mf ) materials . in fact , the rare earth orthochromites have been demonstrated to exhibit ferroelectric polarization either induced by an external magnetic field ( me ) or a spontaneous polarization as a consequence of internal magnetic fields induced by long - range magnetic ordering ( mf ) .
@xcite the potential of the rcro@xmath0 was especially seen in their high magnetic ordering temperatures compared to what was observed in the orthomanganites series , rmno@xmath0 , ( e.g. @xmath6 27 k in tbmno@xmath0 ) which brought about some of the most prominent mf systems . in comparison to the rare earth orthoferrites , rfeo@xmath0 ,
which exhibit even higher magnetic ordering temperatures , the orthochromites , however , are potentially more interesting since the coupling between the rare earth and the transition metal subsystem is larger in magnitude allowing to take advantage of the partially sizeable magnetic moments of the r@xmath5 cations . due to the interaction of the rare earth and 3@xmath7 moments , rcro@xmath0 systems exhibit a variety of complex magnetic structures , often with the presence of a weak ferromagnetic component and several spontaneous spin reorientations.@xcite these were ascribed to the absence of fourth - order crystal field splitting terms of the @xmath8 = 3/2 spin multiplet of the cr@xmath5 cations with an 3@xmath9 electronic configuration.@xcite the rcro@xmath0 compounds crystallize with the orthorhombic gdfeo@xmath0 structure - type ( centrosymmetric spacegroup @xmath10 ) which is highlighted in fig . [ fig1 ] .
bertaut and collaborators analyzed the possible symmetry - compatible ordered magnetic structures of the orthorhombic abo@xmath0 compounds.@xcite with respect to the possible symmetry adapted modes for the cr atoms , four representations of the base vectors are possible , three of them allow a weak ferromagnetic component in the presence of anisotropic exchange forces .
magnetoelectric effects as well as magnetic and electric field induced switching of the dielectric polarization have been detected in smcro@xmath0 , gdcro@xmath0 , and ercro@xmath0.@xcite polarization in ercro@xmath0 disappears below the spin orientation transition of @xmath2 22 k which was ascribed to the crucial connection of the magnetoelectric behavior to the weak ferromagnetism.@xcite this conclusion was subsequently questioned by preethi meher _
_ who observed qualitatively similar results but weaker in magnitude effects in lucro@xmath0 with a diamagnetic r constituent.@xcite at present , there are also diverging opinions about whether the ferroelectric state as evidence by polarization as well as pyroelectric currents occur at the nel temperature as concluded by rajeswaran
_ et al . _ or whether a ferroelectric state is formed already at higher temperatures as proposed earlier by lal _
et al . _ or sahu _
et al._.@xcite for example , prado gonjal _
_ could not identify a strong correlation between the magnetic and dielectric properties of rcro@xmath0 which were prepared by a microwave assisted sintering.@xcite .
orange , red , and gray circles represent europium , oxygen and chromium atoms , respectively .
oxgen atoms connect neighboring cro@xmath11 octahedra via common corners .
structural parameters were taken from the x - ray data which are compiled in table [ table2 ] .
the lattice parameters at room temperature amount to 5.346 , 5.511 , and 7.629 for @xmath12 , @xmath13 , and @xmath14 , respectively .
, height=226 ] the properties of rare earth orthochromites with most of the rare earth elements have been extensively investigated .
it appears that the knowledge of the magnetic properties of the system eucro@xmath0 is limited to magnetization studies and a first raman scattering investigation carried out down to 90 k.@xcite preliminary measurements of the pyroelectric and dielectric properties of eucro@xmath0 have been carried out for @xmath15 300 k.@xcite the magnetization studies found antiferromagnetic ordering below @xmath2175 k and weak ferromagnetism with a small ferromagnetic moment of @xmath20.1 @xmath3 along the @xmath14-axis.@xcite .
the raman data evidenced anomalies in the shift of some raman modes near the nel temperature.@xcite neither the magnetic structure nor a detailed investigation of the magnetoelastic coupling or the thermal properties of eucro@xmath0 have been carried out , so far . in a preceding publication , we described the preparation , characterization and magnetization of eucro@xmath0 and cecro@xmath0 prepared by a solution combustion method and found magnetization irreversibility and exchange bias effects on such samples.@xcite in the present paper , we report on the results of a comprehensive investigation of the magnetic and thermal properties of eucro@xmath0 from room temperature down to liquid helium temperatures .
especially , we performed detailed neutron diffraction experiments using neutrons of different wavelengths and solved the magnetic structure of eucro@xmath0 conclusively .
we confirm weak ferromagnetism due to weak canting of the cr moments .
magnetoelastic effects are investigated by low - temperature thermal expansion and x - ray powder diffraction and detailed raman scattering experiments .
finally , we report on a first characterization of the dielectric properties of eucro@xmath0 by frequency dependent permittivity measurements .
high - purity powder samples of eucro@xmath0 were synthesized by the solution combustion method starting from equimolar solutions of high purity europium nitrate , chromium nitrate and glycine .
the ground powder was first calcinated at 500 @xmath16c for 5 h and reground and sintered at 950 @xmath16c for 12 h. process.the products were characterized with respect to the composition , particle size and phase purity by x - ray diffraction as described in detail elsewhere.@xcite all neutron powder diffraction ( npd ) patterns were collected on 12 g sample powder packed in an al container using the dualspec c2 powder neutron diffractometer , equipped with a curved 800-wire bf@xmath0 detector at the nrc canadian neutron beam center cnbc , canada .
two different wavelengths of 2.37 and 1.3282 wavelengths from the silicon 311 and the 531 reflection , respectively , were used .
a graphite filter was placed in the incident beam in front of the sample to eliminate higher order contributions .
diffraction patterns collected with the short - wavelength neutrons were employed to obtain bragg reflections at high @xmath17 values to enable reliable refinements of nuclear structure whereas the longer wavelength was used to explore the small @xmath17 regime and to search for magnetic scattering .
additional neutron scattering was carried out at the n5 triple axis spectrometer at the nrc canadian neutron beam center to trace the temperature dependence of the order paramater in detail .
the initial and final neutron energies both were chosen by the pyrolytic graphite pg002 reflections . in order to reduce the very high absorption cross section for thermal neutrons from the @xmath18eu isotope in the natural isotope composition of eu , we utilized a flat geometry thin - walled aluminum sample holder which was long enough to cover the full beam height but reduced the thickness of the sample to @xmath2 1 mm.@xcite to trace magnetoelastic coupling effects low - temperature x - ray powder diffraction ( xrpd ) patterns were collected with a bruker d8 discovery system equipped with an oxford instruments closed cycle cooler flat - plate stage .
all profile refinements of the npd and xrpd patterns were performed with the @xmath19 software.@xcite dc magnetization measurements were performed using a magnetic properties measurement system ( quantum design , mpms ) .
heat capacity measurements were done in a physical properties measurement system ( quantum design , ppms ) .
dielectric properties of rectangular shaped samples ( 6.4 @xmath20 1.6 mm@xmath21 ; thickness @xmath2 0.88 mm ) were measured versus temperature at discrete frequencies with an ah2700a capacitance bridge ( andeen - hagerling , inc . ) .
the complex permittivity was analyzed assuming a parallel circuit of a lossless capacitor and a resistor .
the linear thermal expansion of a sample of 0.801(1 ) mm length was measured using a high - resolution miniature dilatometer.@xcite the raman spectra were collected on a jobin yvon typ v 010 labram single grating spectrometer with @xmath21 @xmath22 spectral resolution .
the spectrometer setup included a double super razor edge filter , peltier cooled ccd camera and a mikrocryo cryostat with a copper cold finger .
measurements were performed with linearly polarized he / ne gas laser light of 632.817 nm with @xmath23 1mw of power .
the light beam was focused to a 10 @xmath24 m spot on the top surface of the sample using a microscope .
measurements were taken in temperatures ranging between 10 k and 325 k.
chemical and structural properties and the morphology of the samples used for the subsequent investigation have been described in detail elsewhere.@xcite phase purity of the samples was confirmed using room temperature xrpd .
transmission electron microscopy images revealed a uniform size distribution of the particles ranging between 50 - 70 nm . a random particle size distribution ranging between 60 and 80 nm also allowed to model the observed broadening of the bragg reflections in the xrpd patterns .
[ magneutpatt ] displays an npd pattern of eucro@xmath0 collected with a neutron wavelength of 1.3282 at 3.5 k. an analogous pattern was collected at 280 k ( not shown ) .
both patterns contain dominant reflections from the al sample container which was used to shape a thin sample in order to reduce neutron absorption of the eu isotope @xmath18eu .
in addition to the nuclear scattering of eucro@xmath0 , the 3.5 k npd pattern reveals magnetic scattering with a dominant magnetic bragg reflection at @xmath7 = 4.41(2 ) . for the profile refinement of the npd patterns
, we first refined the bragg reflections of the al container by partially treating them as special reflections allowing individual shifts and widths .
subsequently , the eucro@xmath0 nuclear pattern was refined by varying the lattice parameters , atom positions , and the thermal displacement parameters .
the background was modeled by a tschebyscheff polynomial of higher degree .
the zero - point of the pattern was also refined but it remained below 0.1@xmath25 .
the nuclear structural parameters of the refined npd patterns at 280 and 3.5 k as well as structural parameters obtained from room temperature xrpd patterns , are compiled in table [ table2 ]
. collected at 3.5 k with @xmath26 = 1.3282 .
the black circles represent the measured data , the red solid line is the result of the profile refinement .
the blue solid line at the bottom of the graph shows the difference between the observed and calculated patterns .
the vertical tics ( olive , black , red ) mark the angles of the bragg reflections used to simulate the refined pattern for al , nuclear and magnetic scattering , respectively .
the vertical green arrow shows the ( 011)/(101 ) magnetic reflections near 17.2@xmath25 ( @xmath7 = 4.41(2 ) ) highlighted in the upper left inset.,width=321 ] the positional parameters of all atoms generally agree well with those obtained from the xrpd patterns @xcite , with a slight discrepancy seen for the eu atom positions resulting from the 280 k pattern .
especially , the npd data confirm the atom positional parameters of the oxygen atoms o1 and o2 derived from the xrpd data .
all cell parameters exhibit a decrease by lowering the temperature .
the relative decrease is most pronounced in the @xmath14 lattice parameter .
a shift of the oxygen atoms could not be detected upon lowering the temperature to 3.5 k. furthermore , indication for a structural phase transition between 280 and 3.5 k was not found from the npd patterns .
.structural parameters and conventional reliability indicators of eucro@xmath0 as obtained from the rietveld profile refinement ( _ fullprof _ ) of the xrpd pattern collected with mo-@xmath27 radiation at @xmath28 = 295 k and npd patterns using @xmath26 = 1.3282 at @xmath28 = 3.5 k and 280 k , respectively .
the profile refinements were performed within the space group @xmath10 ( no . 62 ) .
the respective site occupancies were not refined .
an absorption correction of @xmath29 1 cm and 1.5 cm was assumed for the xrpd and the npd refinements , respectively @xcite . if no error bars are given the parameters have been fixed in the refinements . [ cols="^,^,^,^",options="header " , ] our magnetic susceptibility data of eucro@xmath0 show magnetic ordering at about @xmath2 178(1 ) k , as had already been described by tsushima _
_ et al.__@xcite and recently by prado - gonjal _ _ et al.__@xcite the temperature dependence of the magnetization for eucro@xmath0 shown in fig .
[ fig3 ] exhibits a pronounced thermal hysteresis with a splitting of field cooled ( fc ) and zero - field cooled ( zfc ) susceptibility starting at @xmath2178 k , indicating a weak ferromagnetic component with a saturation of the fc component at low temperatures .
this weak ferromagnetic component @xcite has been attributed to a slight spin canting of the cr@xmath5 moments .
.,width=321 ] at sufficiently high temperatures above long - range ordering , the magnetic susceptibility follows a curie - weiss law according to @xmath30 the curie constant , @xmath31 = @xmath32 which depends on the avogadro number @xmath33 , the boltzmann constant @xmath34 , the @xmath35-factor @xmath35 , the bohr magneton @xmath3 and the spin values of the respective magnetic entities is given by @xmath36 the term , @xmath37 = @xmath38 + @xmath39 in eq .
( [ eqcw ] ) , was added to take care of the temperature independent diamagnetic ( @xmath38 ) and van vleck contributions ( @xmath39 ) . from the tabulated diamagnetic increments for the individual ions ,
@xmath38 was estimated to contribute -68@xmath2010@xmath40 @xmath41/mol to the temperature independent part of the magnetic susceptibility.@xcite van vleck contributions which become especially important for the magnetism of eu@xmath5 are weakly temperature dependent and will be discussed below . for convenience sake , one defines the effective magnetic moment as @xmath42 for cr@xmath5 with three electrons occupying the @xmath43 orbitals , in a first approximation , it is justified to assume spin - only magnetism with @xmath44 2 and a spin value of @xmath45 = 3/2.@xcite accordingly , one expects an effective magnetic moment for cr@xmath5close to @xmath46 .
[ invchi ] displays the inverse magnetic susceptibilities of eucro@xmath0 at high temperatures .
the data reveal a magnetic field dependence for small magnetic fields which we attribute to a slight trace of ferromagnetic impurities , possibly euo or cro@xmath47 , which saturates at higher fields .
the data collected at 1 t and 3 t fall on top of each other and these are used for further analysis .
( a ) measured with different applied magnetic fields .
the field dependence arises most likely from a trace ferromagnetic impurity ( e.g. euo or cro@xmath47 , @xmath48 20 ppm ) which is gradually saturated with higher fields .
( b ) inverse magnetic susceptibility of eucro@xmath0 measured with a field of 1 t. the solid red lines represent fits of a curie - weiss law ( eq .
( [ eqcw ] ) ) to the data . for the curie constant we have used the value of 1.873 @xmath41k / mol corresponding to an effective moment of cr@xmath5 of 3.87 @xmath3 .
blue dashed line : the van vleck contribution from eu@xmath5 discussed in detail in the text.,width=321 ] a first analysis of the high temperature susceptibility data in terms of the curie - weiss law ( eq .
( [ eqcw ] ) ) points to a negative curie - weiss temperature about three times larger in magnitude than the long - range ordering temperature .
the negative curie - weiss temperature indicates sizable predominant antiferromagnetic spin exchange interaction .
the derived curie - weiss constant of @xmath31 @xmath48 6.1 @xmath41k / mol gave an effective magnetic moment of 6.98 @xmath49 .
if we attribute the difference to the cr@xmath5 moment simply to the magnetism of the eu@xmath5 ion by using @xmath50 = @xmath51 , we arrive at effective moments for eu@xmath5 of the order of 5.9 @xmath49 , which is by far too large compared to what is expected for eu@xmath5 systems at room temperature ( @xmath52 = 3.4@xmath49 @xcite ) . in order to clarify this discrepancy we followed an alternative approach :
the magnetic moment of eu@xmath5 emerges due to the gradual thermal population of @xmath53 states with @xmath54 taking values of @xmath54 = 0 , 1 , ... , 6 .
accordingly , eu@xmath5 typically shows a weak temperature dependent van vleck paramagnetism with a typical room temperature value of about 4@xmath2010@xmath55 @xmath56/mol.@xcite the effective magnetic moment derived using eq .
( [ mueff ] ) exhibits a strong temperature dependence approaching zero at low temperatures . using the energy separation of the @xmath53 states which is determined by the spin orbit coupling parameter @xmath57 and calculating their thermal population , the effective moment and the susceptibility of eu@xmath58
have been derived e. g. by lueken.@xcite @xmath59(eu@xmath5 ) shows a weak temperature dependence and typically increases from liquid he temperatures to room temperature by about 15%.@xcite in order to fit the magnetic susceptibility of eucro@xmath0 we included the cr part according to @xmath60(cr@xmath5 ) and allowed in eq .
( [ eqcw ] ) in addition to the temperature independent diamagnetic contributions a temperature dependent van vleck susceptibility of eu@xmath5 , @xmath61 .
[ invchi](b ) shows the fit of eq .
( [ eqcw ] ) to the experimental data including diamagnetic , van vleck contribution from eu@xmath5 ions and a curie - weiss contribution from the cr@xmath5 ions applying a curie constant @xmath31 = 1.873 @xmath41k / mol corresponding to an effective moment of @xmath46 . in order to fit the data ,
it was found necessary to multiply the eu contribution by a factor of @xmath21.26 to enhance the van vleck susceptibility .
this enhancement can either be due to the magnetic polarization of the eu moments by the cr moments or it may be ascribed to a slight change of the susceptibility due to crystal field splitting effects of the excited @xmath53 levels which has not taken into consideration in the calculation of the susceptibility of the free eu@xmath5 moments.@xcite as a consequence of the proposed fitting procedure employed to fit the high - temperature magnetic susceptibility of eucro@xmath0 , the curie - weiss temperature is reduced by about @xmath235% but it is found still a factor of two larger than the nel temperature . fig .
[ magnetization ] shows the magnetization versus magnetic field for temperatures above and below the nel temperature of 178 k of a polycrystalline sample of eucro@xmath0 .
below the nel temperature , a remanent moment is striking which amounts to 0.75 @xmath41/g or about 0.033 @xmath3 per cr atom .
if we take a random orientation of the crystallites into consideration , this finding implies a weak ferromagnetic moment per cr atom of @xmath20.1 @xmath3 , in agreement with previous findings.@xcite at various temperatures as indicated in the inset.,width=321 ] the heat capacities of most rare earth orthochromites , rcro@xmath0 ( r = la .
pr , nd , sm .
gd , dy , ho , er and y ) have been measured by satoh _
_ down to liquid nitrogen temperature and the anomalies associated with the long - range magnetic ordering of cr@xmath5 have been well documented.@xcite recently , bartolom _ et al . _ and su _ et al .
_ have extended the heat capacity measurements on the systems ndcro@xmath0 , ybcro@xmath0 , and hocro@xmath0 down to liquid helium temperatures.@xcite it appears that only bartolom _ et al .
_ have ventured to compare the heat capacity with a lattice reference ( lagao@xmath0 ) in order to extract the magnetic contributions to the heat capacity of ndcro@xmath0 .
they were especially interested in the crystal field splitting of the nd@xmath5 ions but also worked out the magnetic entropy associated to the long - range ordering of the cr moments .
to the best of our knowledge , the heat capacity of eucro@xmath0 has not been determined yet .
[ heatcap](a ) shows the heat capacity of eucro@xmath0 and that of lacro@xmath0 and the non - magnetic reference material lagao@xmath0 for comparison .
the latter was taken from the work of bartomolom _
et al._. @xcite the magnetic contribution to the heat capacity of eucro@xmath0 , @xmath62 , ( see inset fig .
[ heatcap](a ) ) was obtained by subtracting a lattice reference which was calculated by down - scaling the temperatures of the heat capacity of lagao@xmath0 by a lindemann factor.@xcite the lindemann factor which takes into account the atom mass differences of the constituents of eucro@xmath0 and lagao@xmath0 was estimated from the atomic weights and the volumes of the nuclear unit cells to @xmath2 0.95.@xcite the magnetic contribution to the heat capacity is characterized by a @xmath26-shaped peak at 176(1 ) k. the anomaly is somewhat narrower than the analogous anomaly found for ndcro@xmath0.@xcite the latter also does not exhibit the shoulder seen in the @xmath62 of eucro@xmath0 at @xmath2 130 k. ( black squares ) , of lacro@xmath0 ( red circles ) , and of lagao@xmath0 ( blue dashed line ) taken from bartolom _ _ et al.__@xcite .
the inset shows the magnetic heat capacity of eucro@xmath0 ( black ) obtained after subtraction of the lattice contribution ( details see text ) and of ndcro@xmath0 taken from ref . ,
( b ) magnetic entropy of eucro@xmath0 , @xmath63 as obtained according to eq .
( [ eqs ] ) .
@xmath64 is the molar gas constant.,width=321 ] the magnetic entropy , @xmath65 , removed by the long - range ordering of the cr moments in eucro@xmath0 was calculated by integrating @xmath66 , according to @xmath67 at 275 k , the magnetic entropy amounts to @xmath68 in agreement with the entropy @xmath69 expected for a @xmath8 = 3/2 system . the detailed temperature dependence of the magnetic entropy displayed in fig .
[ heatcap](b ) reveals that about 70% of the total magnetic entropy are removed below the nel temperature .
only 30% are acquired in the regime of critical fluctuations above @xmath70 .
our heat capacity results confirm long - range magnetic ordering below a nel temperature of 176 k , in agreement with the magnetization data .
the entropy data pinpoint a @xmath8 = 3/2 magnetic system ordering consistent with the @xmath9 configuration of the cr@xmath71 cations while contributions to the magnetic heat capacity arising from the eu@xmath5 cations are not apparent .
_ found a similar result for cr ordering in ndcro@xmath0 , but also detected an additional contribution from nd crystal electric field excitations below @xmath2100 k.@xcite first evidence for magnetic scattering was gained by comparing the diffraction patterns collected at @xmath28 = 280 and 3.5 k on the dualspec diffractometer with neutrons of a wavelength of 1.3282 ( cf .
[ magneutpatt ] ) .
apart from the dominating bragg reflections arising from the al container , weaker nuclear bragg reflections are seen .
in addition to the nuclear bragg peaks , a magnetic bragg reflection was identified at 17.15@xmath25(@xmath7 = 4.41(2 ) ) .
the 280 k pattern is featureless in this part of the diffraction pattern .
measurements at the larger wavelength ( see below ) reveal that this reflection consists of two reflections which can be indexed as ( 011)/(101 ) , the latter one shifted to higher angles , allowing for an understanding of the asymmetric peak shape noticeable in the 1.3282 pattern . the refinement of the magnetic structure will be described in more detail below .
firstly , we will discuss the dependence of the intensity and the splitting of these reflections on the temperature .
[ magref2p37ang ] displays in an overview the temperature dependence of the magnetic reflection observed at @xmath7 = 4.41(2 ) , highlighting its disappearance between 170 and 200 k. a closer inspection of the shape of this reflection reveals a shoulder on the high angle side , indicating that it is composed of two overlapping reflections .
we successfully de - convoluted it by fitting two gaussian functions , with the gaussian at the high angle shoulder having an intensity of about @xmath72 of the main line ( see top right inset in fig .
[ magref2p37ang ] ) .
the splitting of two reflections which amounts to @xmath2 0.7@xmath25 is approximately independent of the temperature , indicating temperature stability of the magnetic structure .
the two reflections can be indexed as ( 011 ) and ( 101 ) .
they are only slightly separated as a consequence of the similar @xmath12 and @xmath13 lattice parameters .
, of the ( 011 ) reflection versus temperature .
the solid red line is a fit to a power law with ( fixed ) critical exponent @xmath73 0.33 and a critical temperature @xmath74 = 178(5 ) k. , width=283 ] a more complete temperature dependence of the ( 011)/(101 ) magnetic reflection was collected on nrc s triple - axis spectrometer n5 , using an elastic configuration by adjusting @xmath75 to 30 mev ( 1.6513 ) and the momentum transfer @xmath17 to 1.4164 @xmath76 .
[ n5powlaw ] displays the integrated intensity versus temperature in a linear and log - log scale versus reduced temperature . measured with neutrons of wavelength 1.6513 versus temperature .
the solid red line represents a fit of a power law according to eq .
( [ eqcrit ] ) with parameters given in the lower inset .
the upper inset displays the integrated intensity in a log - log plot versus the reduced temperature , @xmath77 = ( @xmath28 - @xmath74)/@xmath74 ) , with @xmath74 = 179.4 k. , width=283 ] a power law according to @xmath78 where the integrated intensity @xmath79 @xmath80 , was fitted to the background - corrected intensities by using data within the reduced temperature range @xmath77 = ( @xmath28 - @xmath74)/@xmath74 ) @xmath23 0.25 .
the fits resulted in a critical temperature @xmath81 which is good in agreement with the susceptibility and heat capacity data .
the critical exponent derived from the fit amounts to @xmath82 consistent with the critical exponents of standard universality classes.@xcite the inset in fig .
[ n5powlaw ] shows the data in a log - log scale highlighting the good agreement of the critical power law down to reduced temperatures of 3@xmath2010@xmath55 with a critical exponent given in eq .
( [ critexpo ] ) .
finally , we describe the magnetic structure of eucro@xmath0 .
the symmetry analysis for orthorhombic abo@xmath0 compounds has been carried out and described in detail by bertaut.@xcite according to the symmetry analysis , four representations of the base vectors are possible for the cr atom , occupying the sites ( @xmath83,0,0 ) , ( @xmath83,0,@xmath83 ) ( 0,@xmath83,@xmath83 ) , and ( 0,@xmath83,0 ) .
three of these representations , ( @xmath84 , @xmath84 , and @xmath85 ) , are compatible with a weak ferromagnetic component .
a good fit of magnetic refinement of our npd pattern collected at 3.5 k is achieved based on the @xmath85 representation , i.e. a @xmath1 magnetic structure ( see fig .
[ magstr ] ) with a negligible component along the @xmath86 direction .
the small but finite ferromagnetic component ( @xmath87 ) which was already indicated by a spontaneous magnetization component in the susceptibility data , is discussed below .
the refined magnetic moment of the cr atoms at 3.5 k amounts to @xmath88 consistent with the expected value of 3@xmath3 for a @xmath8= 3/2 system with a @xmath35-factor @xmath89 2 .
[ magstr ] displays the magnetic structure of eucro@xmath0 as refined assuming a dominant @xmath1 spin configuration and a slight canting leading to a weak ferromagnetic moment along @xmath14 of 0.1 @xmath3 . as refined from the npd diffraction data .
the cr moments order with a @xmath1- type antiferromagnetic structure and a slight canting with a @xmath87 component of 0.1 @xmath3 implying a tilting angle of the moments with the @xmath12 - @xmath13 plane of @xmath2 2.4 @xmath90.,width=188 ] in the diffraction pattern collected with @xmath26 = 1.3282 , the ( 011)/(101 ) magnetic reflections are overlapping . choosing a wavelength of 2.37 allows us to resolve them and to compare their intensities with the expected intensities of the @xmath1 magnetic structure .
[ spincanting ] displays the profile refinement as well as a very weak peak at 36.2@xmath25 .
this peak results from the nuclear and magnetic diffraction , the latter is due to the small ferromagnetic component along the @xmath14-axis .
model calculations ( see fig . [ spincanting ] ) indicate the ferromagnetic component to lie between 0.1 and 0.3 @xmath3 , consistent with the magnetization measurements ( see above ) .
this finding implies a canting angle of the cr moment out of the @xmath12-@xmath13 plane between 2.4 and 7@xmath25 . collected at 3.5 k with @xmath26 = 2.37 .
the ( black ) circles represent the measured data , the ( blue , red ) solid lines are the result of the profile refinement ( nuclear part and magnetic part ) with a @xmath1 antiferromagnetic structure and with a weak ferromagnetic component along @xmath14 fixed to 0.1 @xmath3 and 0.3 @xmath3 . the vertical ( black , red ) bars mark the angles of the bragg reflections used to simulate the refined pattern for the nuclear and magnetic scattering , respectively .
the ( blue ) asterisk tags an impurity reflection.,width=283 ] in order to investigate the crystal structure and structural changes below room temperature and to search for magnetoelastic distortions , we have carried out linear thermal expansion , powder xrpd and raman scattering measurements down to liquid helium temperature .
a first overview of thermal and magnetoelastic effects on the lattice properties becomes available from the measurement of the dilatometric length change of a polycrystalline pellet ( fig .
[ thexpa ] ) . starting from lowest temperatures and slowly ( @xmath20.5k / min ) raising the temperature ,
the length change is characterized by an extended plateau whereas the sample starts to contract above @xmath2 70 k , i.e. eucro@xmath0 shows negative thermal expansion between @xmath2 70 and @xmath2 100 k. above @xmath2100k the sample length grows monotonically with a slight change of the slope at @xmath2 175 k , i.e. in the range of the nel temperature . the anomalous behavior below @xmath2100 k is clearly revealed in the coefficient of linear thermal expansion , @xmath91 = @xmath92 ( @xmath93 ) ( see inset fig .
[ thexpa ] ) which is negative between @xmath2 70 and 100 k. , measured by capacitance dilatometry in a warming cycle .
inset : coefficient of the linear thermal expansion , @xmath91.,width=302 ] the thermal expansion measurements have been carried out on a randomly oriented polycrystalline pellet .
detailed directional information of the thermal and magnetic effects on the lattice properties was obtained from a series of powder xrpd measurements over the temperature range from 10 k to room temperature .
[ xrdthexpa ] displays the orthorhombic lattice parameters versus temperature .
while @xmath12 and @xmath14 show about the same temperature behavior , the temperature variation of @xmath13 is by a factor of 2.5 smaller than the contraction of @xmath12 and @xmath14 .
a noticeable anomaly associated with long - range antiferromagnetic ordering below @xmath2 178 k , clearly seen in the thermal expansion measurements is only visible in the @xmath13 lattice parameter .
below @xmath2175 k , @xmath94 exhibits a slight anomalous decrease below the linear part found for all lattice parameters above @xmath2 200 k. @xmath13 also exhibits a slight increase towards lowest temperatures , i.e. a slightly negative thermal expansion whereas @xmath12 and @xmath14 have a positive thermal expansion throughout the whole temperature range .
the cell volume shows positive thermal expansion in the whole temperature range indicating that the small magnetoelastic distortion in @xmath13 is compensated by the temperature behavior of @xmath12 and @xmath14 .
subtle anomalies in all xrpd derived lattice parameters between 50 and 70 k may be associated to the negative thermal expansion anomaly seen in the capacitance dilatometer measurements ( cf .
[ thexpa ] ) . .
the ( red ) solid lines are fits of a linear temperature dependence to the data above 200 k.,width=302 ] raman scattering investigations on the series of orthochromites rcro@xmath0
( @xmath64 = y , la , nd , pr , sm , gd , dy , ho , er , yb , lu ) have been rather extensively carried out at room temperature as well as @xmath2 100 k in order to assign the phonon modes and to study the lattice response to magnetic ordering either of the cr or of the rare earth moments.@xcite our dilatometer measurements and the low - temperature xpd data indicated magnetoelastic distortions of the lattice .
raman spectroscopy of the phonon modes allows to trace very sensitively subtle changes of the lattice and to identify possible structural and magnetic phase transitions . versus temperature .
the mode assignment is given above the topmost ( 300 k ) spectrum.,width=1058 ] at present , only one raman study on eucro@xmath0 down to @xmath290 k is available reporting some anomalies near the nel temperature in the shift of some modes as well as in their linewidth.@xcite here , we have extended these investigation and examined the raman spectra down to 10 k and studied in detail the shift of the modes and their linewidth versus temperature . according to group theory one expects 24 raman active modes for eucro@xmath0 : 7 @xmath95 + 5 @xmath96 + 7 @xmath97 + 5 @xmath98 , involving vibrations of eu and the oxygen atoms.@xcite fig .
[ allraman ] shows the raman spectra between 10 k and room temperature with the assignments according to weber _
et al._.@xcite the spectra contain only 11 raman modes , thus the other predicted modes have either too low an intensity to be observed or they are located below our experimental region .
the spectra are similar to what has been observed for other rare earth orthochromites and from the dependence of the mode energies on the ionic radii of the rare earth atoms fall between the spectra of gdcro@xmath0 and smcro@xmath0 .
the @xmath95 - @xmath96 doublet near 375 @xmath22 is split whereas the mode at @xmath2 470 @xmath22 is still degenerate but clearly broadened as compared to other modes in its neighborhood .
above 600 @xmath22 a broad band is seen which is centered at 650 @xmath22 .
it corresponds to the analogous broad band at @xmath2720 @xmath22 in lacro@xmath0 which has been assigned to lattice imperfections possibly involving cr@xmath99 centers.@xcite at first glance , all raman mode frequencies appear to not be affected by lowering the temperature .
a closer inspection of the shift and the linewidth of the raman modes was carried out by fitting lorentzian lines to the modes .
this procedure discloses especially weak but noticeable softening for all analyzed modes below @xmath2 75 k. some modes show additional magnetic anomalies below @xmath2 175 k. generally , the linewidths decrease and pass through a minimum at @xmath2 75 k and grow again towards lowest temperatures .
noticeable magnetic anomalies could not be detected in the mode linewidths at the nel temperature around 175 k. the temperature dependent change of the frequency of a phonon mode of an insulator with frequency @xmath100 , comprises contributions from the change of the lattice volume , @xmath101 due to thermal expansion or contraction ( including lattice anomalies ) and also magnetostrictive induced changes of the unit cell volume , intrinsic multi - phonon anharmonic contributions , @xmath102 , and contributions from spin - phonon coupling effects , @xmath103 .
these superimpose according to@xcite @xmath104 the lattice contributions are proportional to the volume change , @xmath105 and connected via the mode grneisen parameter , @xmath106 via @xmath107 assuming mode grneisen parameters @xmath108 1 our xpd results indicate mode shifts of the order of 0.4% over the whole temperature range . especially the magnetoelastic effct seen in the lattice parameter
@xmath13 is very weak and can not explain magnetic contributions .
anharmonic decay processes add an additional contribution to the phonon linewidths and lineshifts .
based on the work of klemens assuming three and fourfold phonon - decay processes , balkanski has derived the following relationship for the temperature dependence of the anharmonic line shift and linewidth@xcite @xmath109 where @xmath110 and @xmath111 are bose - einstein statistical factors , @xmath112 and @xmath113 , respectively , with @xmath114 = @xmath115 and @xmath86 = @xmath116 appropriate for the decay of an optical phonon into two and three acoustic phonons of identical energy.@xcite an analogous relationship holds for the phonon linewidth.@xcite spin - phonon coupling caused by the modulation of the spin exchange interaction by lattice vibrations adds another channel for magnetic renormalization of the phonon frequencies.@xcite it is proportional to the spin - spin correlation function @xmath117 and the second derivative of the spin exchange integral @xmath118 with respect to the coordinate connection the spin @xmath119 and @xmath120.@xcite using the molecular field approximation of the spin - spin correlation function , @xmath121 , @xmath122 is proportional to the square of the sublattice magnetization @xmath123 according to @xmath124 spin - phonon coupling renormalization of the mode lineshift is subject to the modulation of the spin exchange and generally effects modes at different frequencies differently , depending on the manner of how the spin exchange is affected by the relevant lattice vibrations modifying the bonding distances and angles especially to the oxygen cations mediating the superexchange .
[ raman160 ] displays the raman shift , the linewidth and the residual @xmath125 - @xmath126 of the @xmath127 at @xmath2160 @xmath22 .
the mode assignment has been adopted from the refs .
. - @xmath126 .
the red solid lines were obtained from a fit of eq .
( [ anharm ] ) to the experimental data taken at temperatures above 200 k and extrapolated to low temperatures .
mode assignment according to refs .. ,width=302 ] raman shift and linewidth above @xmath128 175 k follow well a temperature dependence expected for anharmonic phonon decay processes according to eq .
( [ raman160 ] ) .
anomalies in the raman shift of the 160 @xmath22 @xmath97 mode at @xmath2175 k and near @xmath2 correspond well to antiferromagnetic ordering and to the anomaly seen in the thermal expansion ( see fig .
[ thexpa ] . between 175 and 125 k , the raman shift levels off apparently due to magnetic effects . after a step - like increase below 125 k
, the raman shift passes through a maximum centered at @xmath2 75 k and decreases again to lowest temperatures . in contrast to the raman shift , the linewidth shows no apparent magnetic anomalies in this temperature regime .
though , analogous to the shift , the linewidth attains a minimum at @xmath2 75 k and grows again by about @xmath210% to the lowest temperatures .
[ raman1 ] displays the raman shift of the @xmath95 - @xmath96 doublet near 380 @xmath22 .
in contrast to the @xmath2160 @xmath22 @xmath97 mode , magnetic effects in raman modes at higher frequencies are either weaker or can not be verified due to an increased linewidth of theses modes .
[ ramanfwhm ] shows the lorentzian fwhm of these two modes .
however , the raman shifts as well as the linewidth show a maximum and a minimum centered at @xmath2 75 k similar to the behavior of the 160 @xmath22 @xmath97 mode . - @xmath96 doublet observed near 380 @xmath22 .
mode assignments according to refs .. ,width=302 ] - @xmath96 doublet observed near 380 @xmath22.,width=302 ] similarly , the increase / decrease of the shift / linewidth towards very low temperatures correspond to the negative thermal expansion indicated by the temperature dependence of the lattice parameter @xmath13 and the anomalous length change at these temperatures . at these temperatures ,
magnetic effects of the cr ordering can safely be ruled out since the cr moments have reached saturation at these temperatures ( see figure [ n5powlaw ] ) . to what extend , the eu@xmath5 magnetism and especially the temperature dependent of thermal population of the @xmath129f@xmath130 ( @xmath131 0 ) levels may be responsible for the observed effect remains elusive at present . the structural anomaly detected by the thermal expansion at @xmath128 100 k has a corresponding anomaly in the dielectric properties .
[ dielectric](a ) and ( b ) show the relative permittivity and the dielectric loss @xmath132 of eucro@xmath0 , measured at frequencies of 100 , 1000 and 10000 hz .
on cooling , the relative permittivity ( fig . [ dielectric](a ) ) exhibits a step - like feature which for 100 hz starts at @xmath2 100 k and leads to a decrease of the permittivity by a factor of 2.5 . by increasing the frequency
, the step - like feature moves to higher temperatures .
the permittivity step and its frequency dependence are accompanied by a broad resonance - like feature in dielectric loss @xmath132 in fig .
[ dielectric](b ) .
the peak temperature of the frequency dependent dielectric loss @xmath132 maximum at @xmath133 range between @xmath280 and 125 k. this temperature interval coincides with the position of the highest slope in the capacitance and the observed anomaly in thermal expansion fig .
[ thexpa ] .
measurements in magnetic fields up to 5.5 t did not alter the observed anomalies in dielectric properties . )
( loses ) of polycrystalline sample of eucro@xmath0 for several frequencies.,width=302 ] the frequency shift of @xmath133 exhibits an activated behavior ( fig . [
tmaxactivated ] ) according to @xmath134 with an activation energy , @xmath135 , of @xmath136 and an attempt frequency @xmath137 of @xmath138 .
the ( red ) dashed line represents an arrhenius law according to eq .
( [ tmax ] ) with an activation energy of 170 mev.,width=302 ]
eucro@xmath0 which crystallizes with the gdfeo@xmath0 structure - type orders antiferromagnetically below @xmath2 178 k as we could show using a combination of magnetic susceptibility , heat capacity and npd measurements .
the magnetic susceptibility data reveal weak ferromagnetism below the nel temperature , in agreement with earlier studies.@xcite the magnetic contribution to the heat capacity is characterized by a @xmath26-type anomaly at the nel temperature containing a magnetic entropy in agreement with the entropy expected for a spin @xmath8 = 3/2 magnetic system consistent with the electronic configuration of the cr@xmath5 cations .
our xrpd and npd measurements provide precise atomic positions of all atoms in the cell , especially for the light oxygen atoms .
the low - temperature npd data revealed extra bragg peaks of magnetic origin which can be attributed to a @xmath1 antiferromagnetic structure with an ordered moment of @xmath2 2.4 @xmath3 consistent with an @xmath8 = 3/2 spin - only groundstate .
the weak ferromagnetic moment points along the @xmath14-axis and it amounts to @xmath2 0.1 , consistent with magnetization data .
evidence for an ordering of the eu moments is neither found in the npd data nor in the heat capacity data or magnetization data .
the contribution of the eu@xmath5 cations to the magnetism is rather characterized by a weakly temperature dependent van vleck - type paramagnetism arising from the successive population of @xmath131 0 excited magnetic states .
indication of a polarization of the eu moments by the ordered cr moments can be derived by the fits of the high - temperature ( @xmath139 ) magnetic susceptibility data .
effects of the eu magnetism could be the origin of the slight decrease of the saturation magnetization below 20 k ( see inset fig .
[ fig3 ] ) and the decrease of the raman lineshifts and the linewidth ( see figs .
[ raman160 ] , [ raman1 ] and [ ramanfwhm ] ) and also the slight negative thermal expansion along the @xmath13 axis at low temperatures .
( see fig .
[ xrdthexpa ] ) .
structural anomalies detected in the dielectric raman scattering and thermal expansion data indicate magnetoelastic distortion .
these are clearly visible in the shift of some raman modes at higher temperatures and also the dilatometer measurements of the lenght change carried out on a polycrystalline sample .
most striking is the anomalous contraction of the @xmath13 lattice parameter below @xmath70 .
it can be understood on the basis of the @xmath1-type antiferromagnetic structure with essentially antiparallel arrangement of the nearest neighbor cr moments in the @xmath12 - @xmath13 plane and the tendency to strengthen the antiferromagnetic spin exchange coupling by decreasing the cr - cr distance .
the decrease of the @xmath13 lattice parameter leads to a shortening of cr - o2 distances , i.e. of the cr to the oxygen atoms connecting the cro@xmath11 octahedra in the @xmath12 - @xmath13 plane leading also to a decrease of the nearest neighbor cr - cr distance in the @xmath12 - @xmath13 plane ( 3.84 ) .
such a distance decrease leads to a strengthening of the superexchange interaction.@xcite the effect of the bond - length decrease is partially compensated by a cr - o2 - cr bonding angle reduction .
when decreasing @xmath13 , the bonding angle is slightly moving away from the favoring 180@xmath25 cr - o - cr configuration .
this effect is obviously less determining than that induced by the distance decrease . along the @xmath14-axis ,
the decrease of the @xmath13 parameter also leads to a slight decrease of the cr - o1 bond distance , i.e. of the cr atoms to the apical oxygen atoms in the cro@xmath11 octahedra .
the bonding angle cr - o1 - cr also slightly increases when @xmath13 shrinks .
both effects favor an growth of the antiferromagnetic spin exchange along @xmath14 . in total , the anomalous @xmath13 axis shortening starting below @xmath70 is a consequence of the @xmath1-type antiferromagnetic ordering and the tendency to minimize the exchange energy and to strengthen the antiferromagnetic structure .
we finally discuss the anomalies seen in the dielectric permittivity and the conductance observed at temperatures around @xmath2100 k. the anomalies show activated behavior with an activation energy of 170 mev similar to the activation energy of a relaxor - like ferroelectric features detected in cecro@xmath0 , however at significantly higher temperatures.@xcite for cecro@xmath0 an activation energy of 130 mev was derived .
such energies correspond to typical optical phonon frequencies of oxygen related vibrations suggesting a vibrational origin of the dielectric anomalies . however , the calculated frequencies are largely different from raman oxygen frequencies possibly indicating a different origin of the observed dielectric anomalies in both compounds .
the dielectric anomalies in eucro@xmath0 are paralleled by anomalies in the thermal expansion indicating some origin in lattice anomalies which however could not be clearly detected by the less sensitive xrpd measurements .
the absence of any magnetic field dependence also supports a lattice origin of the dielectric anomalies . in summary , we have conclusively established the cr antiferromagnetic ordering and investigated and discussed associated anomalies in various lattice , thermal and electric properties in eucro@xmath0 .
our investigations add to complete the knowledge of the unusual physical and especially magnetic properties of the rare earth orthochromite series .
this work was financially supported by brock university , the natural sciences and engineering research council of canada ( nserc ) , the ministry of research and innovation ( ontario ) , and the canada foundation for innovation , canada .
we thank e. brcher and g. siegle for expert experimental assistance and j. rodrguez - carvajal for advice with the size / shape refinement of the the particle size . | we carried out detailed studies of the magnetic structure , magnetoelastic coupling , and thermal properties of eucro@xmath0 nano - powders from room temperature to liquid helium temperature .
our neutron powder diffraction and x - ray powder diffraction measurements provide precise atomic positions of all atoms in the cell , especially for the light oxygen atoms .
the low - temperature neutron powder diffraction data revealed extra bragg peaks of magnetic origin which can be attributed to a @xmath1 antiferromagnetic structure with an ordered moment of @xmath2 2.4 @xmath3 consistent with the @xmath4 electronic configuration of the cr@xmath5 cations .
apart from previously reported antiferromagnetic and ferromagnetic transitions in eucro@xmath0 at low temperatures , we also observed an anomaly at about 100 k. this anomaly was observed in temperature dependence of sample s , lattice parameters , thermal expansion , raman spectroscopy , permittivity and conductance measurements .
this anomaly is attributed to the magnetoelastic distortion in the eucro@xmath0 crystal . |
in the past decade , there has been a significant growth in the usage of wireless networks , and in particular , cellular networks , because of the increased data demands .
this has been the driver of recent research for new ways of managing interference in wireless networks . due to the superposition and broadcast properties of the wireless medium ,
interfering signals pose a significant limitation to the rate of communication of users in a wireless network .
hence , it is of interest to understand the fundamental limits of communication in interference channels and to capture the effect of interference on optimal encoding and decoding schemes .
the problem of finding the capacity region of even the simple @xmath11user gaussian interference channel is still an open problem .
however , approximations exist in the literature , where the capacity region or the sum capacity is known in the special scenarios of strong and low interference ( @xcite , @xcite , @xcite , @xcite ) .
another effective approximation that simplifies the problem is to consider only the sum degrees of freedom ( dof ) or the pre - log factor of the sum capacity at high signal - to - noise ratio ( snr ) .
the dof criterion provides an analytically tractable way to characterize the sum capacity and captures the number of interference - free sessions that can be supported in a given multi - user channel . in @xcite
, the dof per user of the fully connected gaussian interference channel was shown to be upper bounded by @xmath12 .
this was shown to be achievable through the interference alignment ( ia ) scheme in @xcite . however , this achievable dof many not be sufficient to meet the demands of wireless applications in many scenarios of practical interest , and hence , it is of interest to study ways to enhance the infrastructure of wireless networks in order to increase the rate of communication .
the considered infrastructural enhancement in this work is the deployment of a backhaul link , through which the transmitters / base stations can exchange messages that they wish to deliver in a cellular downlink session . in order to model the finite capacity of the backhaul link ,
we impose a cooperation constraint where each message can be available at a maximum of @xmath1 transmitters .
we call @xmath1 the _ cooperation order_. the availability of each message at more than one transmitter allows for coordinated multi - point ( comp ) transmission @xcite . in @xcite , a comp transmission model for the fully connected @xmath0-user interference channel was considered .
each message was assumed to be available at the transmitter carrying the same index as the message as well as @xmath13 succeeding transmitters .
using an extension of the asymptotic interference alignment scheme of @xcite , the dof of the channel in this setting was shown to be lower bounded by @xmath14 , and it was conjectured that this lower bound is valid for all values of @xmath0 .
it was then shown that this lower bound is within one degree of freedom of the maximum achievable dof .
we note that this dof cooperation gain beyond @xmath15 does not scale linearly with @xmath0 as @xmath0 goes to infinity .
in other words , the asymptotic per user dof remains @xmath12 . in section [ sec : fc ] , we study whether there exists an assignment of messages satisfying the cooperation order constraint that enables the achievability of an asymptotic per user dof that is strictly greater than @xmath12 .
the assumption of full connectivity is key to the results obtained in @xcite , @xcite , @xcite , and in section [ sec : fc ] of this work . for the fully connected interference channel , interference mitigating schemes
are designed to avoid the interference caused by all other transmitters in the network . however , in practice , each receiver gets most of the destructive interference from a few dominant interfering transmitters . for example , in cellular networks , the number of dominant interfering transmit signals at each receiver ranges from two to seven . all the interference from the remaining transmitters may contribute to the interference floor , and the improvement obtained by including them in the dominant interferers set may not justify the corresponding overhead .
for this reason , we study locally connected channels in section [ sec : lc ] , where the channel coefficients between transmitters and receivers that lie at a distance that is greater than some threshold are approximated to equal zero . for the locally connected channel model
, we assume that each transmitter is connected to @xmath6 neighboring receivers as well as the receiver carrying its own index , @xmath16 preceding receivers and @xmath17 succeeding receivers . the special case of this model where @xmath9 is wyner s asymmetric model @xcite .
this special case was considered in @xcite , and it was assumed that each message is available at the transmitter carrying the same index as well as @xmath13 succeeding transmitters .
the asymptotic per user dof was characterized under this setting as @xmath18 .
the achieving scheme relies only on zero - forcing transmit beamforming . in section
[ sec : lc ] , we extend this result and characterize the asymptotic per user dof for wyner s asymmetric model as @xmath8 under a general cooperation order constraint .
the message assignment enabling this result uses only local cooperation , that is , each message is available only at neighboring transmitters .
the size of the neighborhood does not scale linearly with the size of the network , and therefore , our assignment scheme enjoys the same advantage as the message assignment considered in @xcite .
the remainder of this work is organized as follows .
related work is summarized in section [ sec : relatedwork ] . the problem setup
is then introduced in section [ sec : problemsetup ] .
an informal summary of results is provided in section [ sec : informalsummary ] .
the asymptotic per user dof of the fully connected interference channel with comp transmission is studied in section [ sec : fc ] .
the locally connected channel model is considered in section [ sec : lc ] .
the introduced results are then discussed in section [ sec : discussion ] , and the paper is concluded with some final remarks in section [ sec : conclusions ] .
many existing works studying interference networks with cooperating transmitters use the term _ cognitive radios _
( e.g. @xcite , @xcite , @xcite , @xcite , @xcite ) .
cooperation through cumulative message sharing is studied for the fully connected channel in @xcite , where each message is available at the transmitter carrying the same index and all following transmitters .
we use a similar setting of cooperation to that of cumulative message sharing in the coding scheme for locally connected channels in section [ sec : dofgains ] . in another body of work , unlike the considered setting where we assume that transmitters cooperate by sharing complete messages , cooperation through sharing partial message information that is considered as side information is studied ( see e.g. , @xcite ) . in @xcite and @xcite ,
the transmitters are allowed to cooperate through noise free bit pipes or over the air , respectively .
communication scenarios with cooperating multiple antenna transmitters have been considered in @xcite and @xcite under the umbrella of the x - channel . however , in the x - channel , mutually exclusive parts of each message are given to different transmitters .
this is extended in @xcite to allow each part of each message to be available at more than one transmitter , and in @xcite the mimo x - channel is studied in the setting where transmitters share further side information .
finally , it is worth noting that in the considered setting , we implicitly assume the coordinated design of the transmit beams between all transmitters .
this kind of coordination is also referred to in the literature as _ transmitter cooperation _ , even without the sharing of messages ( see e.g. @xcite ) .
we use the standard model for the @xmath19user interference channel with single - antenna transmitters and receivers , @xmath20 where @xmath21 is the time index , @xmath22 is the transmitted signal of transmitter @xmath23 , @xmath24 is the received signal at receiver @xmath10 , @xmath25 is the zero mean unit variance gaussian noise at receiver @xmath10 , and @xmath26 is the channel coefficient from transmitter @xmath23 to receiver @xmath10 over the time slot @xmath21 . we remove the time index in the rest of the paper for brevity unless it is needed .
we use @xmath27 $ ] to denote the set @xmath28 . for any set
@xmath29 $ ] , we define the complement set @xmath30 , i\notin { \cal a}\}$ ] . for
each @xmath31 $ ] , let @xmath32 be the message intended for receiver @xmath10 .
we use the abbreviations @xmath33 , @xmath34 , @xmath35 , and @xmath36 to denote the sets @xmath37 , @xmath38 , @xmath39 , and @xmath40 , respectively , and the abbreviations @xmath41 , @xmath42 , and @xmath43 to denote the sets @xmath44\right\}$ ] , @xmath45\right\}$ ] , and @xmath46\right\}$ ] , respectively . finally , for @xmath47 $ ] , we let @xmath48 be the @xmath49 matrix of channel coefficients between @xmath50 and @xmath35 . we consider two different channel models in the sequel .
first , we consider a fully connected interference channel where all channel coefficients are drawn independently from a continuous distribution .
we next consider a locally connected channel model where channel coefficients between well separated nodes are approximated to be identically zero .
the locally connected channel model is a function of the number of interferers @xmath6 as follows : @xmath51,\ ] ] and all channel coefficients that are not identically zero are drawn independently from a continuous distribution .
we note that for values of @xmath9 and @xmath52 , the locally connected channel reduces to the commonly known wyner s asymmetric and symmetric linear models , respectively @xcite .
we illustrate examples for the described fully and locally connected channel models in figure [ fig : channelmodel ] .
for each @xmath31 $ ] , let @xmath53 $ ] be the transmit set of receiver @xmath10 , i.e. , those transmitters with the knowledge of @xmath32 .
the transmitters in @xmath54 cooperatively transmit the message @xmath32 to the receiver @xmath10 .
the messages @xmath55 are assumed to be independent of each other .
the _ cooperation order _
@xmath1 is defined to be the maximum transmit set size , @xmath56 for any set @xmath29 $ ] , we define @xmath57 as the set of messages carried by transmitters with indices in @xmath58 , i.e. , the set @xmath59 .
let @xmath60 be the average transmit power constraint at each transmitter , and let @xmath61 denote the alphabet for message @xmath32 .
then the rates @xmath62 are achievable if the decoding error probabilities of all messages can be simultaneously made arbitrarily small for large enough @xmath63 , and this holds for almost all channel realizations .
the degrees of freedom @xmath64,$ ] are defined as @xmath65 .
the dof region @xmath66 is the closure of the set of all achievable dof tuples .
the total number of degrees of freedom ( @xmath67 ) is the maximum value of the sum of the achievable degrees of freedom , @xmath68 } d_i$ ] . for a @xmath0-user channel ,
we define @xmath69 as the best achievable @xmath67 over all choices of transmit sets satisfying the cooperation order constraint in .
similarly , we define @xmath70 for a locally connected channel with @xmath6 interfering signals per receiver . in order to simplify our analysis
, we define the asymptotic per user dof @xmath71 , and @xmath72 to measure how @xmath69 , and @xmath70 scale with @xmath0 , respectively , while all other parameters are fixed , @xmath73 @xmath74 for the locally connected channel model where @xmath75 , let @xmath76 .
we silence the first @xmath77 transmitters , deactivate the last @xmath77 receivers , and relabel the transmit signals to obtain a ( @xmath78)-user channel , where the transmitter @xmath10 is connected to receivers in the set @xmath79 .
we note that the new channel model gives the same value of @xmath72 as the original one , since @xmath80 . unless explicitly stated otherwise , we will be using this equivalent model in the rest of the paper .
we show an example construction of the equivalent channel model in figure [ fig : locallyconnected ] .
a message assignment strategy is defined by a sequence of transmit sets @xmath81 , k\in\{1,2,\ldots\}$ ] . for each positive integer @xmath0 and @xmath82 $ ] , @xmath83 , |{\cal t}_{i , k}| \leq m$ ] .
we use message assignment strategies to define the transmit sets for a sequence of @xmath19user channels .
the @xmath84 channel in the sequence has @xmath85 users , and the transmit sets for this channel are defined as follows .
the transmit set of receiver @xmath10 in the @xmath84 channel in the sequence is the transmit set @xmath86 of the message assignment strategy .
we call a message assignment strategy _ optimal _ for a sequence of @xmath19user fully connected channels , @xmath87 , if and only if there exists a sequence of coding schemes achieving @xmath71 using the transmit sets defined by the message assignment strategy
. a similar definition applies for locally connected channels .
we say that a message assignment strategy satisfies the local cooperation constraint , if and only if there exists a function @xmath88 such that @xmath89 , and @xmath90 , \forall k\in { \bm z}^{+}.\ ] ] let @xmath91 and @xmath92 be the maximum achievable asymptotic per user dof @xmath71 and @xmath72 under the additional local cooperation constraint , respectively .
in this paper , we study the benefit of comp transmission via the asymptotic per user dof .
in particular , we investigate whether the asymptotic per user dof increases by allowing comp transmission , i.e. , by allowing a cooperation order @xmath93 , and characterize this improvement , if it exists , as a function of @xmath1 .
the considered problem is completely described by two system parameters , namely , the channel connectivity and the cooperation order @xmath1 .
we attempt to find an answer by setting two design parameters : the message assignment strategy satisfying the cooperation order constraint , and the achievable scheme .
we know from the results in @xcite and @xcite that the asymptotic per user dof of the fully connected channel is @xmath2 if each message is available only at the transmitter carrying the same index ; it is straightforward to extend this result to the case where each message can be available at any single transmitter , i.e. , the case where @xmath94 , and hence , we know that @xmath95 . in @xcite , it was shown that comp transmission achieves a dof gain for the fully connected channel . however , this gain does not scale linearly with the number of users @xmath0 .
the considered message assignment strategy in @xcite is the spiral strategy where each message is assigned to the transmitter carrying the same index as well as @xmath13 succeeding transmitters .
we note that this strategy satisfies the local cooperation constraint defined in section [ sec : localcooperation ] , and show in section [ sec : asymptoticpudof ] that local cooperation can not achieve an asymptotic per user dof gain for the fully connected channel .
more precisely , we show that @xmath96 furthermore , we extend this negative conclusion in section [ sec : asymptoticpudof ] to all message assignments that are restricted to assign each message to at most two transmitters , i.e. , @xmath97 in general , we prove in theorem [ thm : tauouterbound ] the following upper bound on the asymptotic per user dof for any value of @xmath1 , @xmath98 we then show that the tightness of for the case where @xmath99 does not generalize .
in particular , we prove in theorem [ thm : taufthree ] the following tighter bound for the case where @xmath100 , @xmath101 we summarize the results obtained for the fully connected channel model in table [ tab : fc ] .
.summary of results for fully connected channels . [ cols="<,<,<,<,<",options="header " , ] in section [ sec : fcupperbound ] , we restate from @xcite a necessary condition on any point in the dof region of the fully connected channel with a fixed message assignment .
we then use this condition to derive corollary [ cor : dofouterbound ] that implies directly all of the provided dof upper bounds for the fully connected channel .
moreover , corollary [ cor : dofouterbound ] can be used to shed insight on the open problem of determining whether @xmath102 for @xmath103 . in @xcite
, we showed that if corollary [ cor : dofouterbound ] is tight for all instances of the problem , then scalable dof cooperation gains are achievable for @xmath104 , i.e. , it would follow that @xmath102 for @xmath103 .
it is obvious for locally connected channels that some message assignments are _
reducible_. for example , for the case where @xmath94 , any assignment of a message @xmath32 to a transmitter that is not connected to the @xmath105 receiver can not achieve a positive rate for communication of that message . to prove dof upper bounds for locally connected channels with comp transmission
, we use a characterization of necessary conditions on _ irreducible message assignments _ , as discussed in section [ sec : lcusefulmsgassignment ] .
in this section , we investigate whether @xmath106 for @xmath93 , and message assignment strategies that may lead to a positive conclusion . we know from @xcite , and @xcite that the per user dof of a fully connected interference channel without cooperation is @xmath2 , i.e. , @xmath107 . in @xcite ,
the following spiral message assignment strategy was considered for @xmath108 : @xmath109,\\ \{i , i+1,\ldots , k,1,2,\ldots , m-(k - i+1)\},\quad & \forall i \in \{k-(m-2),k-(m-2)+1,\ldots , k\ } , \end{cases}$ ] using this message assignment strategy and an asymptotic interference alignment scheme , it was shown in ( * ? ? ?
* theorem @xmath110 ) that @xmath111 and it was shown in @xcite that this lower bound is within one degree of freedom from the maximum achievable dof using the spiral message assignment strategy .
however , even if @xmath112 for all values of @xmath0 , the dof gain due to comp transmission ( beyond @xmath113 ) does not scale with the number of users @xmath0 . hence , the question of whether @xmath106 for @xmath114 remains open
here , we note that the spiral message assignment strategy satisfies the local cooperation constraint and in section [ sec : asymptoticpudof ] , we generalize the negative conclusion of @xcite to all message assignment strategies satisfying the local cooperation constraint . in order to characterize the dof of the channel @xmath71
, we need to consider all possible strategies for message assignments satisfying the cooperation order constraint defined in . through the following corollary of ( * ? ? ?
* theorem @xmath3 ) , we provide a way to upper bound the maximum achievable dof for each such assignment , thereby , introducing a criterion for comparing different message assignments satisfying using the special cases where this bound holds tightly .
we first restate the following result from @xcite .
[ @xcite ] [ thm : dofouterbound ] any point @xmath115 in the dof region of a @xmath19user fully connected channel satisfies the inequalities : @xmath116 .
\end{split}\ ] ] the proof is availble in ( * ? ? ?
* theorem @xmath3 ) .
we now provide the following corollary for bounding the dof number @xmath67 of a @xmath19user fully connected channel with a fixed message assignment .
recall that for a set of transmitter indices @xmath117 , the set @xmath118 is the set of messages carried by transmitters in @xmath117 .
[ cor : dofouterbound ] for any @xmath119 , if there exists a set @xmath117 of indices for transmitters carrying no more than @xmath120 messages , and @xmath121 , then @xmath122 , or more precisely , @xmath123}\max ( |c_{\cal s}| , k-|{\cal s}|).\ ] ] for each subset of transmitter indices @xmath124 $ ] , we apply lemma [ thm : dofouterbound ] with the sets @xmath58 and @xmath125 assigned as follows , @xmath126 and @xmath127 .
we refer the reader to figure [ fig : dofouterbound ] for an example illustration of corollary [ cor : dofouterbound ] . , with @xmath128 and @xmath129 .
transmit signals with indices in @xmath117 , and messages as well as receive signals with indices in @xmath118 are shown in tilted red font and dashed boxes . the dof @xmath130 . ]
we now use corollary [ cor : dofouterbound ] to prove upper bounds on the asymptotic per user dof @xmath71 . in an attempt to reduce the complexity of the problem of finding an optimal message assignment strategy
, we begin by considering message assignment strategies satisfying the local cooperation constraint defined in section [ sec : localcooperation ] .
we now show that a scalable cooperation dof gain can not be achieved using local cooperation .
[ thm : localcooperation ] any message assignment strategy satisfying the local cooperation constraint of can not be used to achieve an asymptotic per user dof greater than that achieved without cooperation .
more precisely , @xmath131 fix @xmath132 . for any value of @xmath133
, we use corollary [ cor : dofouterbound ] with the set @xmath134 . note that @xmath135 , and hence , it follows that @xmath136 . finally , @xmath137 .
the lower bound follows from @xcite without cooperation .
we now investigate if it is possible for the cooperation gain to scale linearly with @xmath0 for fixed @xmath1 .
it was shown in theorem [ thm : localcooperation ] that such a gain is not possible for message assignment strategies that satisfy the local cooperation constraint . here
, we only impose the cooperation order constraint in and prove in theorem [ thm : tauouterbound ] an upper bound on @xmath71 that is tight enough for finding @xmath138 .
[ thm : tauouterbound ] for any cooperation order constraint @xmath5 , the following upper bound holds for the asymptotic per user dof , @xmath139 for any value of @xmath1 and @xmath0 , we show that @xmath140 . for every value of @xmath0
such that @xmath141 is an integer , we show that @xmath142 . when @xmath141 is not an integer , we add @xmath143 extra users such that @xmath144 is an integer , and bound the dof as follows , @xmath145 it then suffice to consider the case where @xmath141 is an integer
the idea is to show that for any assignment of messages satisfying the cooperation order constraint , there exists a set of indices @xmath124 $ ] for @xmath141 transmitters that do not carry more than @xmath146 messages , and then the dof upper bound follows by applying corollary [ cor : dofouterbound ] .
more precisely , it suffices to show that the following holds , @xmath147 : |{\cal s}|=\frac{k-1}{m } , |{\cal c}_{\cal s}|=\frac{k(m-1)+1}{m}=k-|{\cal s}|.\ ] ] we first illustrate simple examples that demonstrate the validity of .
consider the case where @xmath148 , @xmath99 , we need to show in this case that there exists a transmitter that does not carry more than two messages , which follows by the pigeonhole principle since each message can only be available at a maximum of two transmitters .
now , consider the slightly more complex example of @xmath149 , @xmath99 , we need to show in this case that there exists a set of two transmitters that do not carry more than three messages . we know that there is a transmitter carrying at most two messages , and we select this transmitter as the first element of the desired set . without loss of generality , let the two messages available at the selected transmitter be @xmath150 and @xmath151 .
now , we need to find another transmitter that carries at most one message among the messages in the set @xmath152 .
since each of these three messages can be available at a maximum of two transmitters , and we have four transmitters to choose from , one of these transmitters has to carry at most one of these messages . by adding the transmitter satisfying this condition as the second element of the set , we obtain a set of two transmitters carrying no more than three messages , and holds .
we extend the argument used in the above examples through lemmas [ lem : basis ] and [ lem : inductionstep ] in the appendix .
we know by induction using these lemmas that holds , and the theorem statement follows .
together with the achievability result in @xcite , the statement in theorem [ thm : tauouterbound ] implies the following corollary .
[ cor : tauftwo ] for any message assignment strategy such that each message is available at a maximum of two transmitters , the asymptotic per user dof is the same as that achieved without cooperation .
more precisely , @xmath153 the characterization of @xmath71 for values of @xmath154 remains an open question , as theorem [ thm : tauouterbound ] is only an upper bound .
moreover , the following result shows that the upper bound in theorem [ thm : tauouterbound ] is loose for @xmath100 .
[ thm : taufthree ] for any message assignment strategy such that each message is available at a maximum of three transmitters , the following bound holds for the asymptotic per user dof , @xmath155 in a similar fashion to the proof of theorem [ thm : tauouterbound ] , we prove the statement by induction . the idea is to prove the existence of a set @xmath117 with approximately @xmath156 transmitter indices , and these transmitters are carrying no more than approximately @xmath157 messages , and then use corollary [ cor : dofouterbound ] to derive the dof outer bound . in the proof of theorem [ thm : tauouterbound ] , we used lemmas [ lem : basis ] and [ lem : inductionstep ] in the appendix , to provide the basis and induction step of the proof , respectively . here , we follow the same path until we show that there exists a set @xmath117 such that @xmath158 and @xmath159 , and then we use lemma [ lem : mthreeinductionstep ] in the appendix to provide a stronger induction step that establishes a tighter bound on the size of the set @xmath118 .
we note that it suffices to show that @xmath160 for all values of @xmath0 such that @xmath161 is an even positive integer , and hence , we make that assumption for @xmath0 .
define the following , @xmath162 @xmath163 @xmath164 now , we note that , @xmath165 and by induction , it follows from lemmas [ lem : basis ] and [ lem : inductionstep ] that @xmath166 $ ] , @xmath167 , @xmath168 .
we now apply induction again with the set @xmath169 as a basis , and use lemma [ lem : mthreeinductionstep ] for the induction step to show that @xmath170 $ ] , @xmath171 , @xmath172 .
hence , we get the following upper bound using corollary [ cor : dofouterbound ] , @xmath173 from which holds .
we note that all the dof upper bounding proofs used so far employ corollary [ cor : dofouterbound ] . in @xcite
, we showed that under the hypothesis that the upper bound in corollary [ cor : dofouterbound ] is tight for any @xmath0-user fully connected interference channel with a cooperation order constraint @xmath1 , then scalable dof cooperation gains are achievable for any value of @xmath104 .
hence , a solution to the general problem necessitates the discovery of either new upper bounding techniques or new coding schemes .
however , as we show next , scalable dof cooperation gains are possible when the assumption of full connectivity is relaxed .
in section [ sec : channelmodel ] , we defined the locally connected channel model as a function of the number of dominant interferers per receiver @xmath6 , by connecting each transmitter to @xmath16 preceding receivers and @xmath17 succeeding receivers , and in section [ sec : dofmodel ] , we illustrated an equivalent model in terms of the asymptotic per user dof @xmath72 . in the equivalent model , each transmitter
is connected to @xmath6 succeeding receivers .
more precisely , we consider the following channel model , @xmath174,\ ] ] and all non - zero channel coefficients are generic . in @xcite , the special case of wyner s asymmetric model ( @xmath9 ) was considered , and the spiral message assignment strategy mentioned in section [ sec : fcpriorwork ] was fixed , i.e. , each message is assigned to its own transmitter as well as @xmath13 following transmitters .
the asymptotic per user dof was then characterized as @xmath18 .
this shows for our problem that , @xmath175 in ( * ? ? ?
* remark @xmath4 ) , a message assignment strategy was described to enable the achievability of an asymptotic per user dof as high as @xmath176 , it can be easily verified that this is indeed true , and hence , we know that , @xmath177 the main difference in the strategy described in ( * ? ? ?
* remark @xmath4 ) from the spiral message assignment strategy considered in @xcite , is that unlike the spiral strategy , messages are assigned to transmitters in an asymmetric fashion , where we say that a message assignment is symmetric if and only if for all @xmath178 , j > i$ ] , the transmit set @xmath179 is obtained by shifting forward the indices of the elements of the transmit set @xmath54 by @xmath180 .
we show that both the message assignment startegy analyzed in @xcite and the one suggested in @xcite are suboptimal for @xmath9 , and the value of @xmath181 is in fact strictly higher than the bounds in and .
the key idea enabling our result is that each message need not be available at the transmitter carrying its own index .
we start by illustrating a simple example for the case of no cooperation ( @xmath94 ) that highlights the idea behind our scheme . per user dof for @xmath182 .
each transmitter is carrying a message for the receiver connected to it by a solid line .
figure showing only signals corresponding to the first @xmath183 users in a general @xmath19user network .
signals in dashed boxes are deactivated .
note that the deactivation of @xmath184 splits this part of the network from the rest . ]
let @xmath150 be available at the first transmitter , @xmath185 be available at the second transmitter , and deactivate both the second receiver and the third transmitter .
then it is easily seen that messages @xmath150 and @xmath185 can be received without interfering signals at their corresponding receivers .
moreover , the deactivation of @xmath184 splits this part of the network from the rest .
i.e. , the same scheme can be repeated by assigning @xmath186 , to the transmitters with transmit signals @xmath187 , respectively , and so on .
thus , @xmath4 degrees of freedom can be achieved for each set of @xmath183 users , thereby , achieving an asymptotic per user dof of @xmath188 .
the described message assignment is depicted in figure [ fig : examplemone ] .
it is evident now that a constraint that is only a function of the load on the _ backhaul _ link may lead to a discovery of better message assignments than the one considered in @xcite . in the following section ,
we show that the optimal message assignment strategy under the cooperation order constraint is different from the spiral strategy .
the above described message assignment strategy and coding scheme for the special case of @xmath182 are shown to be optimal .
it is worth noting that the optimality of a tdma scheme in this case follows as a special case from a general result in @xcite , where necessary and sufficient conditions on channel connectivity and message assignment are derived for tdma schemes to be optimal . in this section ,
we specialize the scheme introduced for multiple - antenna transmitters in ( * ? ? ?
* section iv ) to our setting .
we consider a simple linear precoding coding scheme , where each message is assigned to a set of transmitters with successive indices , and a zero - forcing transmit beamforming strategy is employed .
the transmit signal at the @xmath189 transmitter is given by , @xmath190 where @xmath191 depends only on message @xmath32 .
using simple zero - forcing transmit beams with a fractional reuse scheme that activates only a subset of transmitters and receivers in each channel use , we extend the example in section [ sec : lcexample ] to achieve scalable dof cooperation gains for any value of @xmath192 .
[ thm : lowerbound ] the following lower bound holds for the asymptotic per user dof of a locally connected channel with connectivity parameter @xmath6 , @xmath193 showing that @xmath194 follows by a straightforward extension of the asymptotic interference alignment scheme of @xcite , and hence , it suffices to show that @xmath195 .
we treat the network as a set of clusters , each consisting of consecutive @xmath196 transceivers .
the last @xmath6 transmitters of each cluster are deactivated to eliminate _
inter - cluster _ interference .
it then suffices to show that @xmath197 dof can be achieved in each cluster . without loss of generality , consider the cluster with users of indices in the set @xmath198 $ ] .
we define the following subsets of @xmath198 $ ] , @xmath199 , \\{\cal s}_2 & = & \{l+m+1,l+m+2,\ldots , l+2m\}.\end{aligned}\ ] ] we next show that each user in @xmath200 achieves one degree of freedom , while messages @xmath201 are not transmitted . in the proposed scheme ,
users in the set @xmath169 are served by transmitters in the set @xmath202 and users in the set @xmath203 are served by transmitters in the set @xmath204 .
let the message assignments be as follows .
+ @xmath205 + now , we note that messages with indices in @xmath169 are not available outside transmitters with indices in @xmath206 $ ] , and hence , do not cause interference at receivers with indices in @xmath203 .
also , messages with indices in @xmath203 are not available at transmitters with indices in @xmath206 $ ] , and hence , do not cause interference at receivers with indices in @xmath169 . in order to complete the proof by showing that each user in @xmath200 achieves one degree of freedom ,
we next show that transmissions corresponding to messages with indices in @xmath169(@xmath203 ) do not cause interference at receivers with indices in the same set . to avoid redundancy ,
we only describe in detail the design of transmit beams for message @xmath150 to cancel its interference at all receivers in @xmath169 except its own receiver .
first , the encoding of @xmath150 into @xmath207 at the first transmitter is done in a way that is oblivious to the existence of other receivers in the network except the first receiver , and a capacity achieving code for the point - to - point link @xmath208 is used .
we then design @xmath209 at the second transmitter to cancel the interference caused by @xmath150 at the second receiver , i.e. , @xmath210 similarly , the transmit beam @xmath211 is then designed to cancel the interference caused by @xmath150 at the third receiver .
the transmit beams @xmath212 are successively designed with respect to order of the index @xmath10 such that the received signal due to @xmath213 at the @xmath105 receiver cancels the interference caused by @xmath150 . in general , the availability of channel state information at the transmitters allows a design for the transmit beams for message @xmath32 that delivers it to the @xmath105 receiver with a capacity achieving point - to - point code and simultaneously cancels its effect at receivers with indices in the set @xmath214 , where , + @xmath215 + note that both @xmath216 and @xmath217 equal the empty set , because both @xmath218 and @xmath219 do not contribute to interfering signals at receivers with indices in the set @xmath200 .
we conclude that each receiver with index in the set @xmath220 suffers only from gaussian noise , thereby enjoying one degree of freedom .
we refer the reader to figure [ fig : codingscheme ] for an illustration of the above described coding scheme .
we note that in the above coding scheme , some messages are not being transmitted in order to allow for interference - free communication for the remaining messages .
it is worth noting that this can be done while maintaining fairness in the allocation of the available dof over all users through _ fractional reuse _ in a system where multiple sessions of communication take place , and different sets of receivers are deactivated in different sessions , e.g. , in different time slots or different sub - carriers ( in an ofdm system ) . in order to find an upper bound on the per user dof @xmath72
, we have to consider all possible message assignment strategies satisfying the cooperation order constraint . in this section ,
we characterize necessary conditions for the optimal message assignment .
the constraints we provide for transmit sets are governed by the connectivity pattern of the channel .
for example ,
for the case where @xmath94 , any assignment of message @xmath32 to a transmitter that is not connected to @xmath221 is _ reducible _ , i.e. , the rate of transmitting message @xmath32 has to be zero for these assignments , and hence , removing @xmath32 from its carrying transmitter does not reduce the sum rate in these cases .
we now introduce a graph theoretic representation that simplifies the presentation of the necessary conditions on irreducible message assignments . for message @xmath32 , and a fixed transmit set @xmath54
, we construct the following graph @xmath222 that has @xmath27 $ ] as its set of vertices , and an edge exists between any given pair of vertices @xmath223 $ ] if and only if : * @xmath224 , * @xmath225 .
vertices corresponding to transmitters connected to @xmath221 are given a special mark , i.e. , vertices with labels in the set @xmath226 are marked for the considered channel model .
we now have the following statement .
[ lem : usefulmsgdistribution ] for any @xmath227 such that the vertex @xmath85 in @xmath222 is not connected to a marked vertex , removing @xmath85 from @xmath54 does not decrease the sum rate .
let @xmath117 denote the set of indices of vertices in a component with no marked vertices .
we need to show that removing any transmitter index in @xmath117 from @xmath54 does not decrease the sum rate .
let @xmath228 be the set of indices of received signals that are connected to at least one transmitter with an index in @xmath117 . to prove the lemma , we consider two scenarios , where we add a _ tilde _ over symbols denoting signals belonging to the second scenario . for the first scenario ,
@xmath32 is made available at all transmitters with indices in @xmath117 .
let @xmath229 be a random variable that is independent of all messages and has the same distribution as @xmath32 , then for the second scenario , @xmath32 is not available at any transmitter with an index in @xmath117 , and a realization @xmath230 of @xmath229 is generated and given to all transmitters with indices in @xmath117 before communication starts .
moreover , the given realization @xmath231 contributes to the encoding of @xmath232 in the same fashion as a message @xmath233 contributes to @xmath234 . assuming a reliable communication scheme for the first scenario that uses a large block length @xmath63
, the following argument shows that the achievable sum rate is also achievable after removing @xmath32 from the designated transmitters . and therefore , proving that removing any transmitter in @xmath117 from @xmath54 does not decrease the sum rate .
@xmath235 where @xmath236 is the entropy function for discrete random variables , @xmath237 follows from fano s inequality , @xmath238 follows as the difference between the two scenarios lies in the encoding of @xmath234 which affects only @xmath239 , and @xmath240 holds because any two transmitters carrying @xmath32 and connected to a receiver whose index is in @xmath241 must belong to the same component , and hence , @xmath32 contributes to @xmath242 only through @xmath234 , it follows that @xmath243 has the same joint distribution as @xmath244 for every @xmath245 .
now , it follows that , @xmath246 and hence , the rates @xmath247 $ ] are achievable in the second scenario . in a @xmath248user channel with @xmath9 .
marked vertices are represented with filled circles .
@xmath185 can be removed at both @xmath249 and @xmath250 without decreasing the sum rate , as the corresponding vertices lie in a component that does not contain a marked vertex . ]
we call a message assignment _ irreducible _ if no element in it can be removed without decreasing the sum rate .
the following corollary to the above lemma characterizes a necessary condition for any message assignment satisfying the cooperation order constraint in to be irreducible .
recall that two vertices in a graph @xmath251 are at a distance @xmath252 if and only if the shortest path in @xmath251 between the two vertices has @xmath252 edges . [ cor : usefulmsgdistribution ]
let @xmath54 be an irreducible message assignment and @xmath253 , then @xmath254 , k\in{\cal t}_i$ ] only if the vertex @xmath85 in @xmath222 lies at a distance that is less than or equal @xmath13 from a marked vertex .
note that in the considered channel model , the above result implies that @xmath255 , from which we obtain the following result .
local cooperation is optimal for locally connected channels , @xmath256 and so we note that even though local cooperation does not achieve a scalable dof gain for the fully connected channel , not only does it achieve a scalable gain when the connectivity assumption is relaxed to local connectivity , but the confinement to local cooperation no longer results in a loss in the available dof . in this section ,
we prove upper bounds on @xmath181 and @xmath257 that establishes the tightness of the lower bound in theorem [ thm : lowerbound ] for the special cases where either @xmath9 or @xmath94 .
first , in order to assess the optimality of the coding scheme introduced in section [ sec : dofgains ] for arbitrary values of the system parameters , we prove a general upper bound for a class of coding schemes that only employs a zero - forcing transmit beam - forming strategy .
consider only coding schemes with transmit signals of the form and each message is either not transmitted or allocated one degree of freedom .
more precisely , let @xmath258 $ ] , then in addition to the constraint in , it is either case that the mutual information @xmath259 or it is the case that @xmath260 completely determines @xmath261 . note that @xmath261 can be determined from @xmath260 for the case where user @xmath23 enjoys interference - free communication and @xmath262 for the other case where @xmath260 is not transmitted .
we say that the @xmath189 receiver is _ active _ if and only if @xmath263 .
note that using zero - forcing transmit beamforming , if the @xmath189 receiver is active , then @xmath264 .
let @xmath265 denote the asymptotic characterization of the per user dof under the restriction to the above described class of coding schemes . in theorem
[ thm : zfupperbound ] below , we show that the coding scheme in the proof of theorem [ thm : lowerbound ] achieves the optimal value of @xmath265 .
we first prove lemma [ lem : lemone ] that bounds the number of receivers at which the interference of a given message can be cancelled . for a set
@xmath266 $ ] , let @xmath267 be the set of indices for active receivers connected to transmitters with indices in @xmath117 .
more precisely , @xmath268 , where @xmath269 is the empty set . to obtain the following results
, we assume that for each transmitter in @xmath54 , message @xmath32 contributes to the transmit signal of this transmitter .
i.e. , @xmath270 .
note that this assumption does not introduce a loss in generality , because otherwise the transmitter can be removed from @xmath54 .
we need the following lemma for the proof of the upper bound on @xmath265 in theorem [ thm : zfupperbound ] .
[ lem : lemone ] for any message @xmath32 , the number of active receivers connected to at least one transmitter carrying the message is no greater than the number of transmitters carrying the message . @xmath271 . for each receiver @xmath272
, there exists a transmit signal @xmath273 , @xmath274 $ ] such that conditioned on all other transmit signals , the received signal @xmath275 is correlated with the message @xmath32 .
more precisely , @xmath276 , v\neq k\}\right)>0 $ ] .
now , since we impose the constraint @xmath277 , the interference seen at all receivers in @xmath278 has to be cancelled . finally ,
since the probability of a zero lebesgue measure set of channel realizations is zero , the @xmath279 transmit signals carrying @xmath32 can not be designed to cancel @xmath32 at more than @xmath280 receivers for almost all channel realizations .
[ thm : zfupperbound ] under the restriction to zf transmit beam - forming coding schemes ( interference avoidance ) , the asymptotic per user dof of a locally connected channel with connectivity parameter @xmath6 is given by , @xmath281 the proof of the lower bound is the same as the proof of theorem [ thm : lowerbound ] for the case where @xmath282 .
it then suffices to show that @xmath283 . in order to prove the upper bound ,
we show that the sum degree of freedom in each set @xmath124 $ ] of consecutive @xmath196 users is bounded by @xmath197 .
we now focus on proving this statement by fixing a set @xmath117 of consecutive @xmath196 users , and make the following definitions . for a user
@xmath284 $ ] , let @xmath285 be the set of active users in @xmath117 with an index @xmath286 , i.e. , @xmath287 similarly , let @xmath288 be the set of active users in @xmath117 with an index @xmath289 , @xmath290 assume that @xmath117 has at least @xmath291 active users , then there is an active user in @xmath117 that lies in the middle of a subset of @xmath291 active users in @xmath117 .
more precisely , @xmath292 , we let this middle user have the @xmath105 index for the rest of the proof .
let @xmath293 and @xmath294 be the users in @xmath117 with minimum and maximum indices , respectively , i.e. , @xmath295 and @xmath296 , we then consider the following cases to complete the proof , * case 1 : * @xmath32 is being transmitted from a transmitter that is connected to the receiver with index @xmath293 , i.e. , @xmath297 .
it follows from lemma [ lem : usefulmsgdistribution ] that @xmath298 , and hence , @xmath299 , which contradicts , as @xmath300 .
* case 2 : * @xmath32 is being transmitted from a transmitter that is connected to the receiver with index @xmath294 , i.e. , @xmath301 .
it follows from lemma [ lem : usefulmsgdistribution ] that @xmath302 , and hence , @xmath299 , which again contradicts .
* case 3 : * for the remaining case , there is no transmitter in @xmath54 that is connected to any of the receivers with indices @xmath293 and @xmath294 . in this case , it follows from lemma [ lem : usefulmsgdistribution ] that @xmath54 does not contain a transmitter that is connected to a receiver with an index less than @xmath293 or greater than @xmath294 , and hence , all the receivers connected to transmitters carrying @xmath32 belong to @xmath117 .
it follows that at least @xmath303 receivers in @xmath117 are connected to one or more transmitter in @xmath54 , and since @xmath117 has at least @xmath291 active receivers , then any subset of @xmath303 receivers in @xmath117 has to have at least @xmath304 active receivers , and the statement is proved by reaching a contradiction to in the last case .
now , we consider the special case of @xmath9 , and prove that the lower bound stated in theorem [ thm : lowerbound ] is tight in this case .
we start by stating the following auxiliary lemma for any @xmath0-user gaussian interference channel with a dof number of @xmath67 . for any set of receiver indices
@xmath29 $ ] , define @xmath305 as the set of indices of transmitters that exclusively carry the messages for the receivers in @xmath58 , and the complement set @xmath306 is the set of indices of transmitters that carry messages for receivers outside @xmath58 .
more precisely , @xmath307\backslash\cup_{i \notin { \cal a } } { \cal t}_i$ ] , then , [ lem : dofouterbound ] if there exists a set @xmath308 $ ] , a function @xmath309 , and a function @xmath310 whose definition does not depend on the transmit power constraint @xmath60 , and @xmath311 , then @xmath312 .
the proof is available in the appendix . here
, we provide a sketch .
recall that @xmath313 , and @xmath314 , and note that @xmath315 is the set of transmit signals that do not carry messages outside @xmath33 . fix a reliable communication scheme for the considered @xmath19user channel , and assume that there is only one centralized decoder that has access to the received signals @xmath35 .
we show that using the centralized decoder , the only uncertainty in recovering all the messages @xmath316}$ ] is due to the gaussian noise signals . in this case , the sum dof is bounded by @xmath317 , as it is the number of received signals used for decoding . using @xmath35 ,
the messages @xmath33 can be recovered reliably , and hence , the signals @xmath318 can be reconstructed . using @xmath35 and @xmath318 ,
the remaining transmit signals can be approximately reconstructed using the function @xmath309 of the hypothesis . finally ,
using all transmit signals , the received signals @xmath319 can be approximately reconstructed , and the messages @xmath320 can then be recovered .
we note that lemma [ lem : dofouterbound ] applies to all considered channel models .
now , we use it to prove a dof upper bound for wyner s model .
[ thm : asymmetric model ] the asymptotic per user dof for wyner s asymmetric model with comp transmission is given by , @xmath321 the lower bound follows from theorem [ thm : lowerbound ] . in order to prove the converse , we use lemma [ lem : dofouterbound ] with a set @xmath58 of size @xmath322 . we also prove the upper bound for the channel after removing the first @xmath1 transmitters @xmath323}\right)$ ] , while noting that this will be a valid bound on @xmath181 since the number of removed transmitters is @xmath324 . inspired by the coding scheme in the proof of theorem [ thm : lowerbound ]
, we define the set @xmath58 as the set of receivers that are _ active _ in the coding scheme .
i.e. , the complement set @xmath325 , i= ( 2m+1)(j-1)+m+1 , j \in { \bm{z}^+}\}$ ] . we know from corollary [ cor : usefulmsgdistribution ] that messages belonging to the set @xmath320 do not contribute to transmit signals with indices that are multiples of @xmath291 , i.e. , @xmath326 for all @xmath327 $ ] that is a multiple of @xmath291 . more precisely , let the set @xmath117 be defined as follows : @xmath328 , i \text { is a multiple of } 2m+1\},\ ] ] then @xmath329 . in particular , @xmath330 , and hence it suffices using lemma [ lem : dofouterbound ] to show the existence of linear functions @xmath309 and @xmath310 such that @xmath331}+f_2(z_{\cal a})$ ] . in what follows
we show how to reconstruct a noisy version of the signals in the set @xmath204
@xmath332 @xmath333 , where the reconstruction noise depends only on @xmath36 in a linear fashion
. then it will be clear by symmetry how to reconstruct the rest of transmit signals in the set @xmath334}$ ] .
since @xmath335 and @xmath336 is also given , @xmath337 can be reconstructed .
now , with the knowledge of @xmath337 and @xmath338 , we can reconstruct @xmath339 , and so by iterative processing , a noisy version of all transmit signals in the set @xmath204 can be reconstructed , where the noise is a linear function of the signals @xmath340 . in a similar fashion , given @xmath341 and @xmath342 , the signal @xmath343 can be reconstructed . then with the knowledge of @xmath344 , we can reconstruct @xmath345 , and we can proceed along this path to reconstruct a noisy version of all transmit signals in the set @xmath333 , where the noise is a linear function of the signals @xmath346 .
this proves the existence of linear functions @xmath309 and @xmath310 such that @xmath347}+f_2(z_{\cal a})$ ] , and the coefficients for @xmath310 do not depend on the transmit power constraint @xmath60 , and so by lemma [ lem : dofouterbound ] we obtain the converse of theorem [ thm : asymmetric model ] . in figure
[ fig : mthree ] ( b ) , we illustrate how the proof works for the case where @xmath100 .
note that the missing received signals @xmath348 in the upper bound proof correspond to the inactive receivers in the coding scheme .
we note that even for the case of _ no cooperation _ , an asymptotic per user dof of more than @xmath2 per user dof is achievable , i.e. , @xmath349 . also , it is straightforward to see that the interference alignment scheme can be generalized to show that @xmath350 for any locally connected channel with parameter @xmath6 .
the next theorem generalizes the upper bound in @xcite for locally connected channels , where each message can be available at one transmitter that is not necessarily the transmitter carrying its own index . in particular
, we show that @xmath351 only if @xmath352 .
the following lemma serves as a building block for the upper bound proof in theorem [ thm : nocooperation ] below .
we define @xmath353 as the set of indices of received signals that are connected to transmitter @xmath354 , i.e. , @xmath355 .
note that as we are considering the case of _ no cooperation _
, hence , @xmath54 contains only one element . recall that @xmath356 denotes the available dof for the communication of message @xmath32 .
[ lem : cornerstone ] if @xmath357 , then @xmath358 .
we assume that all messages other than @xmath32 and @xmath359 are deterministic , and then apply lemma [ lem : dofouterbound ] with the set @xmath360 , and functions @xmath309 and @xmath310 defined such that the following holds , @xmath361\backslash\{j\}}\right)&= & h_{s , j}^{-1}\left(y_s-{\bm h}_{\{s\},[k]\backslash\{j\}}x_{[k]\backslash\{j\}}\right ) \\&= & x_j + h_{s , j}^{-1}z_s \\&= & x_j + f_2(z_s),\end{aligned}\ ] ] and then the bound follows .
[ thm : nocooperation ] without cooperation ( @xmath94 ) , the asymptotic per user dof of locally connected channels is given by , @xmath362 the case where @xmath9 is a special case of the result in theorem [ thm : asymmetric model ] .
the lower bound for the case where @xmath363 follows by assigning each message to the transmitter with the same index , and a simple extension of the asymptotic interference alignment scheme of @xcite , and hence , it suffices to show that , @xmath364 in this proof , we use the locally connected channel model defined in .
each transmitter is connected to @xmath16 preceding receivers and @xmath17 succeeding receivers . in order to prove the theorem statement
, we establish the stronger statement , @xmath365 we prove by induction
. the basis to the induction step is given by the following .
@xmath366 the proof of follows from lemma [ lem : cornerstone ] and the fact that all transmitters connected to @xmath367 are also connected to @xmath368 . in order to state the induction step , we first define @xmath369 as a boolean variable that is true if and only if the following is true : * @xmath370 .
* @xmath371 .
the induction step is given by the following .
@xmath372 in order to prove , consider the assignment of message @xmath373 , and note that @xmath373 is available at a transmitter connected to @xmath374 . now , note that @xmath375 , the channel model of implies that any transmitter connected to @xmath374 is either connected to @xmath376 , or to both @xmath377 and @xmath378 .
the proof follows by considering these two cases separately .
* case 1 : * if @xmath373 is available at a transmitter that is connected to @xmath376 , then it follows from lemma [ lem : cornerstone ] that @xmath379 .
since @xmath369 is true , it follows that @xmath370 , and hence , @xmath380 . in this case , holds since @xmath381 is true .
* case 2 : * if @xmath373 is available at a transmitter that is connected to both @xmath377 and @xmath378 , then it follows from lemma [ lem : cornerstone ] that @xmath382 , and @xmath383 .
now , since @xmath369 is true , it follows that @xmath384 , and hence , @xmath385 also , since @xmath369 is true , we know that @xmath386 , and hence , we get from that @xmath387 . in this case , holds since @xmath388 is true .
it follows by induction from and that it is either the case that @xmath389 is true , or @xmath390 is true . if @xmath389 is true , then @xmath391 , and the dof number @xmath392 .
if @xmath393 is true , then it follows that the dof number @xmath394 .
there are two design parameters in the considered problem , the message assignment strategy satisfying the cooperation order constraint , and the design of transmit beams .
we characterized the asymptotic per user dof when one of the design parameters is restricted to a special choice , i.e. , restricting message assignment strategies by a local cooperation constraint or restricting the design of transmit beams to zero - forcing transmit beams .
the restriction of one of the design parameters can significantly simplify the problem because of the inter - dependence of the two design parameters .
on one hand , the achievable scheme is enabled by the choice of the message assignment strategy , and on the other hand , the assignment of messages to transmitters is governed by the technique followed in the design of transmit beams , e.g. zero - forcing transmit beamforming or interference alignment . in the following ,
we discuss each of the design parameters .
the assignment of each message to more than one transmitter ( comp transmission ) creates a virtual multiple input single output ( miso ) network .
a real miso network , where multiple dedicated antennas are assigned to the transmission of each message , differs from the created virtual one in two aspects .
first , in a comp transmission setting , the same transmit antenna can carry more than one message .
second , for locally connected channels , the number of receivers at which a message causes undesired interference depends on the number of transmit antennas carrying the message . for fully connected channels , the number of receivers at which a message causes undesired interference is the same regardless of the size of the transmit set as long as it is non - empty .
the only aspect that governs the assignment of messages to transmitters is the pattern of overlap between transmit sets corresponding to different messages .
it is expected that the larger the sizes of the intersections between sets of messages carried by different transmit antennas , the more dependent the coefficients of the virtual miso channel are , and hence , the lower the available dof . for the spiral assignments of messages considered in @xcite , @xmath395 , and the same value holds for the size of the intersection between sets of messages carried by successive transmitters . in general ,
local cooperation implies large intersections between sets of messages carried by different transmitters , and hence , the negative conclusion we reached for @xmath396 . for the case where we are restricted to zero - forcing transmit beamforming as in section [ sec : dofgains ] , the number of receivers at which each message causes undesired interference governs the choice of transmit sets , and hence , we saw that for locally connected channels , the message assignment strategy illustrated in theorem [ thm : lowerbound ] selects transmit sets that consist of successive transmitters , to minimize the number of receivers at which each message should be cancelled .
this strategy is optimal under the restriction to zero - forcing transmit beamforming schemes .
while it was shown in @xcite that comp transmission accompanied by both zero - forcing transmit beams and asymptotic interference alignment can achieve a dof cooperation gain beyond what can be achieved using only transmit zero - forcing , this is not obvious for locally connected channels . unlike in the fully connected channel , the addition of a transmitter to a transmit set in a locally connected channel may result in an increase in the number of receivers at which the message causes undesired interference .
we note that unlike asymptotic interference alignment scheme , the zero - forcing transmit beamforming scheme illustrated in section [ sec : dofgains ] does not need symbol extensions , since it achieves the stated dof of theorem [ thm : lowerbound ] in one channel realization .
however , it is not clear whether asymptotic interference alignment can be used to show an asymptotic per user dof cooperation gain beyond that achieved through simple zero - forcing transmit beamforming ; we believe that the answer to this question is closely related to both problems that remain open after this work , i.e. , characterizing @xmath71 and @xmath72 .
we studied the dof gain achieved through comp transmission . in particular
, it was of interest to know whether the achievable gain scales linearly with @xmath0 as it goes to infinity , under a cooperation constraint that only limits the number of transmitters at which any message can be available by a cooperation order @xmath1 .
we showed that the answer is negative for the fully connected channel where message assignment strategies satisfy the local cooperation constraint , as well as all possible message assignments for the case where @xmath99 .
the problem is still open for fully connected channels and values of @xmath104 . for locally connected channels where each transmitter is connected to the receiver carrying the same index as well as @xmath6 neighboring receivers , we showed that the asymptotic per user dof is lower bounded by @xmath7 .
the achieving coding scheme is simple as it relies only on zero - forcing transmit beamforming .
we showed that this lower bound is tight for the case where @xmath9 .
in particular , the characterized asymptotic per user dof for that case is @xmath8 , and is higher than previous results in @xcite , and @xcite .
we also revealed insights on the optimal way of assigning messages to transmitters under a cooperation order constraint .
for instance , we considered a local cooperation constraint , where each message can only be available at a neighborhood of transmitters whose size does not scale linearly with the number of users .
while we showed that local cooperation does not achieve a scalable dof gain for the fully connected channel , we also showed that local cooperation is optimal for locally connected channels .
furthermore , we have shed light on the intimate relation between the selection of message assignments and the design of transmit beams .
we have shown that assigning messages to successive transmitters is beneficial for zero - forcing transmit beamforming in locally connected channels as it minimizes the number of receivers at which each message causes undesired interference .
however , the same message assignment strategy can be an impediment to other techniques such as asymptotic interference alignment , because the overlap of sets of messages carried by transmit antennas is large for this assignment of messages .
the authors thank dr .
michle wigger and dr .
shlomo shamai ( shitz ) for interesting discussions related to this paper .
[ lem : basis ] @xmath397 \text { such that } |c_{\{i\}}| \leq m.\ ] ] the statement follows by the pigeonhole principle , since the following holds , @xmath398 [ lem : inductionstep ] for @xmath399 , if @xmath400 $ ] such that @xmath401 , and @xmath402 , then @xmath403 $ ] such that @xmath404 , and @xmath405 .
we only consider the case where @xmath406 , as otherwise , the statement trivially holds . in this case
, we can show that @xmath407 the proof of is available in lemma [ lem : genineq ] below .
note that the left hand side in the above equation is the maximum number of message instances for messages outside the set @xmath57 , i.e. , @xmath408 , i \notin { \cal a } } |c_{\{i\ } } \backslash c_{\cal a}| & \leq & m(k-|c_{\cal a}|)\nonumber \\&<&(k - n)((m-1)(n+1)+2-|c_{\cal a}|).\nonumber \\\end{aligned}\ ] ] since the number of transmitters outside the set @xmath58 is @xmath409 , it follows by the pigeonhole principle that there exists a transmitter whose index is outside @xmath58 and carries at most @xmath410 messages whose indices are outside @xmath57 .
more precisely , @xmath411\backslash{\cal a } : |c_{\{i\}}\backslash{c_{\cal a}}| \leq ( m-1)(n+1)+1-|c_{\cal a}|.\ ] ] it follows that there exists a transmitter whose index is outside the set @xmath58 and can be added to the set @xmath58 to form the set @xmath125 that satisfies the statement .
[ lem : genineq ] if @xmath412 , @xmath5 , and @xmath413 $ ] such that @xmath414 , then the following holds , @xmath415 we first prove the statement for the case where @xmath416 .
this directly follows as , @xmath417 in order to complete the proof , we note that each decrement of @xmath418 leads to an increase in the left hand side by @xmath1 , and in the right hand side by @xmath409 , and , @xmath419 [ lem : mthreeinductionstep ] for @xmath100 , if @xmath400 $ ] such that @xmath420 , and @xmath421 , @xmath422 , then @xmath423 $ ] such that @xmath404 , @xmath424 .
the proof follows in a similar fashion to that of lemma [ lem : inductionstep ] .
let @xmath425 .
we only consider the case where @xmath426 , as otherwise , the proof is trivial .
we first assume the following , @xmath427 now , it follows that , @xmath408 , i \notin { \cal a } } |c_{\{i\ } } \backslash c_{\cal a}| & \leq & m(k-|c_{\cal a}|)\nonumber \\&<&(k - n)\left(n+\frac{k+1}{4}+3-|c_{\cal a}|\right),\nonumber \\\end{aligned}\ ] ] and hence , @xmath411\backslash{\cal a } : |c_{\{i\}}\backslash{c_{\cal a}}| \leq n+\frac{k+1}{4}+2-|c_{\cal a}|,\ ] ] and then the set @xmath428 satisfies the statement of the lemma . finally , we need to show that is true . for the case
where @xmath429 , @xmath430 and hence , @xmath431 , which implies for the case where @xmath429 .
moreover , we note that each decrement of @xmath432 increases the left hand side of by @xmath183 and the right hand side by @xmath433 , and we know that , @xmath434 and hence , @xmath435 , so there is no loss of generality in assuming that @xmath429 in the proof of , and the proof is complete .
in order to prove the lemma , we show that using a reliable communication scheme with the aid of a signal that is within @xmath436 , all the messages can be recovered from the set of received signals @xmath35 .
it follows that any achievable degree of freedom for the channel is also achievable for another channel that has only those receivers , thus proving the upper bound . in any reliable @xmath63-block coding scheme , @xmath437.\ ] ]
therefore , @xmath438 now , the sum @xmath439 } r_i = \sum_{i \in \bar{{\cal a } } } r_i + \sum_{i \in { \cal a}}r_i$ ] can be bounded as @xmath440 where @xmath441 can be made arbitrarily small , by choosing @xmath63 large enough .
the two terms on the right hand side of can be bounded as @xmath442 @xmath443 now , we have @xmath444 therefore , if we show that @xmath445 then from , we have the required outer bound .
since @xmath33 contains all the messages carried by transmitters with indices @xmath305 , they determine @xmath446
. therefore , @xmath447 where @xmath237 follows from the existence of the function @xmath309 such that @xmath448 .
recall that for @xmath449,{\cal s}_2\subseteq[k]$ ] , @xmath450 denotes the @xmath451 matrix of channel coefficients between @xmath452 and @xmath453 , then @xmath238 follows as , @xmath454 and hence , @xmath455 , v. s. annapureddy , and v. v. veervalli , `` on optimal message assignments for interference channel with comp transmission , '' in _ proc .
@xmath456 annual conference on information sciences and systems _ ,
princeton , nj , mar .
, v. s. annapureddy , and v. v. veervalli , `` degrees of freedom ( dof ) of locally connected interference channels with coordinated multi - point ( comp ) transmission , '' _ in proc .
ieee international conference on communications ( icc ) _ , ottawa , jun .
2012 .
x. shang , g. kramer , and b. chen `` a new outer bound and noisy - interference sum - rate capacity for the gaussian interference channel , '' _ ieee trans .
inf . theory _ ,
55 , no . 2 , pp .
689 699 , feb . 2009 .
v. s. annapureddy and v. v. veeravalli `` gaussian interference networks : sum capacity in the low interference regime and new outerbounds on the capacity region , '' _ ieee trans .
inf . theory _ ,
55 , no . 6 , pp . 3032 3050 , jun .
2009 .
v. s. annapureddy , a. el gamal , and v. v. veervalli , `` degrees of freedom of interference channels with comp transmission and reception , '' _ ieee trans .
inf . theory _
58 , no . 9 , pp . 5740 - 5760 , sep .
2012 .
a. lapidoth , n. levy , s. shamai ( shitz ) and m. a. wigger `` cognitive wyner networks with clustered decoding , '' _ submitted to ieee trans .
theory , available at http://arxiv.org/abs/1203.3659._ mar .
2012 m. maddah - ali , a. motahari , and a. khandani `` communication over mimo x channels : interference alignment , decompositions , and performance analysis , '' _ ieee trans .
inf . theory _ ,
54 , no . 8 , pp . 3457 3470 , 2008 .
v. s. annapureddy , a. el gamal , and v. v. veervalli , `` degrees of freedom of cooperative interference networks , '' in _ proc .
ieee international symposium on information theory ( isit ) _ , saint petersburg , aug .
2011 .
s. shamai ( shitz ) and m. a. wigger , `` rate - limited transmitter - cooperation in wyner s asymmetric interference network , '' in _ proc .
ieee international symposium on information theory ( isit ) _ , saint petersburg , aug .
2011 .
, v. s. annapureddy , and v. v. veervalli , `` degrees of freedom ( dof ) of locally connected interference channels with cooperating multiple - antenna transmitters , '' _ in proc .
ieee international symposium on information theory _ , cambridge , ma , jul | coordinated multi - point ( comp ) transmission is an infrastructural enhancement under consideration for next generation wireless networks . in this work ,
the capacity gain achieved through comp transmission is studied in various models of wireless networks that have practical significance .
the capacity gain is analyzed through the degrees of freedom ( dof ) criterion .
the dof available for communication provides an analytically tractable way to characterize the capacity of interference channels .
the considered channel model has @xmath0 transmitter / receiver pairs , and each receiver is interested in one unique message from a set of @xmath0 independent messages .
each message can be available at more than one transmitter .
the maximum number of transmitters at which each message can be available , is defined as the _ cooperation order _ @xmath1 .
for fully connected interference channels , it is shown that the asymptotic per user dof , as @xmath0 goes to infinity , remains at @xmath2 as @xmath1 is increased from @xmath3 to @xmath4 .
furthermore , the same negative result is shown to hold for all @xmath5 for any message assignment that satisfies a _ local cooperation _ constraint .
on the other hand , when the assumption of full connectivity is relaxed to _ local connectivity _ , and each transmitter is connected only to its own receiver as well as @xmath6 neighboring receivers , it is shown that local cooperation is optimal . the asymptotic per user dof is shown to be at least @xmath7 for locally connected channels , and is shown to be @xmath8 for the special case of _ wyner s asymmetric model _ where @xmath9 .
an interesting feature of the proposed achievability scheme is that it relies on simple zero - forcing transmit beams and does not require symbol extensions .
also , to achieve the optimal per user dof for wyner s model , messages are assigned to transmitters in an asymmetric fashion unlike traditional assignments where message @xmath10 has to be available at transmitter @xmath10 .
it is also worth noting that some receivers have to be inactive , and _ fractional reuse _ is needed to achieve equal dof for all users . |
we perform event - driven brownian dynamics ( edbd ) simulations @xcite of @xmath2 self - propelled hard spheres with a gaussian size polydispersity of @xmath3 and a mean diameter of @xmath4 .
the self - propulsion force for particle @xmath5 is modelled as a constant force @xmath6 with an orientation @xmath7 , which undergoes free brownian rotation @xcite .
we employ @xmath8 as our unit of velocity in order to have a direct comparison of the self - propulsion force to the thermal brownian motion . to relate to experiments
, one can employ the stokes - einstein relation with @xmath9 and @xmath10 the viscosity of the solvent .
note that the self - propulsion speed of a single particle is given by @xmath11 . to prepare the initial configuration
, we use the lubachevsky - stillinger algorithm @xcite to grow the particles in the simulation box to the packing fraction of interest .
subsequently , we `` equilibrate '' the system until each particle has moved on average more than several times its own diameter , and the system has reached a steady state .
we confirm that there is no crystalline order observed in all simulations . to investigate the dynamical relaxation of the system
, we measure the self - intermediate scattering function @xmath12 \right\ } \right \rangle}$ ] , where @xmath13 is the wave vector for which we use @xmath14 .
figure [ fig1 ] shows the structural relaxation time @xmath15 , defined by @xmath16 , as a function of packing fraction @xmath17 for self - propulsion @xmath18 ( passive spheres ) , 10 , 20 , 80 , and 800 .
we clearly observe that the dynamics slows down with orders of magnitude upon increasing the packing fraction for both passive and active systems .
in addition , we find that the structural relaxation time of both passive and active systems can be fitted with the vogel - fulcher - tammann ( vft ) law , @xmath19 and the mode coupling theory ( mct ) prediction , @xmath20 @xcite . for a passive hard - sphere system , we find in agreement with refs .
@xcite that the structural relaxation @xmath15 can be described by an algebraic mct divergence at a packing fraction @xmath21 , and an exponential vft divergence at @xmath22 . for active hard spheres
, we observe that @xmath15 decreases by orders of magnitude upon increasing the self - propulsion @xmath6 at fixed @xmath17 .
however , for sufficiently high densities the dynamics slows down dramatically and the structural relaxation time is again well - described by the mct and vft predictions as shown in fig .
we thus find that the dynamical relaxation behavior of active hard spheres is very similar to that of passive hard spheres , as our results show clear evidence for dynamical arrest and a diverging structural relaxation time even for systems that are far out - of - equilibrium at very high activity . in table
[ tab1 ] , we present the resulting critical packing fractions @xmath23 and @xmath24 as obtained from the vft and mct fits , respectively , as well as the critical exponent @xmath25 and prefactor @xmath26 .
we find that @xmath24 as obtained from the mct fits increases monotonically with self - propulsion @xmath6 . for passive hard spheres
, we find an mct critical exponent @xmath27 , which agrees with previous reported values of @xmath28 $ ] @xcite .
in addition , we note that the critical exponent @xmath25 decreases slightly with increasing @xmath6 , but we also mention here that @xmath25 is extremely sensitive to the precise @xmath17-range that is used in the mct fitting as also noted in refs .
@xcite . similarly , by fitting the structural relaxation time @xmath29 with the vft prediction , we find that @xmath30 also increases monotonically with self - propulsion @xmath6 , but less pronounced than @xmath24 .
we also note that @xmath23 reaches a constant value @xmath31 within our statistical accuracy for @xmath32 .
figure [ fig2 ] shows @xmath23 and @xmath24 as a function of the inverse activity @xmath33 .
we find that the packing fraction range @xmath34 for which the dynamical relaxation is characterized by an exponential vft divergence decreases with decreasing @xmath33 , but remains finite even in the limit of infinite activity , i.e. , @xmath35 .
this also suggests that even with infinitely strong random self - propulsions , during the process of producing rcp , the system yet undergoes an mct glass transition , in which the structural relaxation time diverges with a power law , and the obtained rcp of hard spheres still depends on the speed of compression .
we also observe that the prefactor @xmath26 as determined from the vft fits decreases with activity @xmath6 , which corresponds to a stronger dependence of @xmath29 with packing fraction @xmath17 .
the @xmath17-dependence of @xmath29 allows us to classify the various glasses in terms of their `` fragility '' .
the concept of `` fragility '' was originally introduced for molecular glasses to describe the sensitivity of @xmath29 to changes in temperature : fragile glasses show a strong temperature - dependence of @xmath29 , whereas strong glass - formers correspond to a weak sensitivity to temperature .
the analogy between volume - dependent fragility in colloidal glasses and the temperature - dependent fragility in molecular glasses , where @xmath36 is exchanged with @xmath37 , has been demonstrated recently in experiments on deformable colloidal particles , where softer and more deformable particles lead to stronger glass behavior compared to hard particles @xcite .
we thus find from our vft fits that @xmath29 depends stronger on packing fraction @xmath17 upon increasing @xmath6 as the prefactor @xmath26 decreases , which corresponds to more fragile glass behavior .
the fragility is related to the structural relaxation in glasses , which can be described by activated processes like cage - breaking or barrier hopping in a complex energy landscape that reflects the possible glass configurations . upon increasing packing fraction @xmath17 ,
the barriers between local minima in the energy landscape that correspond to different glass configurations increase and the system gets kinetically trapped in a local minimum at short times , and jumps only to other minima on a longer time scale due to activated dynamics . assuming that the relaxation time is determined by the barrier height of the activated process
, one can determine the effective activation energy ( or entropic barrier height ) @xmath38 from the vft fits using @xmath39 .
in the case of passive hard spheres , we find that @xmath40 at @xmath41 , which agrees well with the theoretical predictions of ref .
however , the entropic barrier @xmath38 for passive hard spheres as obtained from the vft fits increases rapidly with @xmath17 , and reaches a value of @xmath42 at @xmath43 , which is much higher than the theoretically predicted barrier height of @xmath44 in ref .
@xcite , and @xmath45 as obtained from the experimental observations and analysis of the structural rearrangements in colloidal glasses under shear @xcite .
a possible explanation for the sharp increase in relaxation time @xmath29 and barrier height @xmath38 with increasing packing fraction @xmath17 is that a larger cooperativity of the particles is required in order to cross the barrier . for self - propelled spheres , we find that the effective activation energy @xmath38 at @xmath46 decreases to only a few @xmath47 for self - propulsions @xmath48 .
hence , the particles can easily escape out of their cages due to the activated dynamics of the self - propulsions , which is consistent with the fact that @xmath46 is below the mct critical packing fraction @xmath24 . for active hard spheres ,
we observe that the relaxation time decreases by orders of magnitude as it becomes much easier for self - propelling motorized hard spheres to cross the barriers and to explore different glass configurations than for passive spheres . to better understand the effect of self - propulsion on the dynamics of hard - sphere glasses , we plot the self - intermediate scattering function @xmath49 for a system of active hard spheres with varying self - propulsions @xmath6 at @xmath50 in fig . [ fig3]a .
we observe that @xmath49 reaches a cage - trapping plateau and remains stuck in an amorphous configuration in the case of passive hard spheres , whereas for active spheres @xmath49 decays to zero within our simulations , and the relaxation time decreases significantly with self - propulsion .
the common picture of glassy dynamics and barrier crossings is that a particle rattles for a long time inside a cage that is formed by its neighbours and then suddenly breaks out of its cage . to investigate whether or not the barrier crossing corresponds to an abrupt cage - breaking event and
how localized these events are , we measure the four - point dynamic susceptibility , @xmath51 , where @xmath52 denotes the fluctuating part of @xmath49 , as shown in fig . [ fig3]a .
as expected , @xmath53 increases initially in time and shows a peak on a time scale that coincides with the structural relaxation time @xmath15 before it decays at longer times .
the height of the peak of @xmath53 is related to the average number of particles that are dynamically correlated in the structural relaxation process @xcite .
figure [ fig3]a shows that the peak height of @xmath53 is @xmath54 for @xmath55 , providing support that there is almost no collective motion involved in the structural relaxation , which is to be expected as @xmath56 is far below the mct critical packing fraction @xmath24 at @xmath55 .
in addition , we find clearly that the peak height increases upon decreasing the self - propulsion towards @xmath57 and thus also decreasing mct critical packing fraction @xmath24 .
we note however that we were not able to determine the peak height at @xmath57 as the relaxation time exceeds our simulation time .
we thus find that even for self - propelled hard spheres the structural relaxation becomes more cooperative for packing fractions close to the mct glass transition , i.e. , @xmath58 . to corroborate our finding on the cooperative nature of the structural relaxation for self - propelled spheres , we show in fig .
[ fig4 ] the distribution of displacements in a system of active hard spheres at @xmath50 measured over a time window that spans the structural relaxation time @xmath15 .
figure [ fig4 ] clearly shows that for @xmath59 the fast moving particles , i.e. , particles with displacements larger than @xmath60 , are spatially correlated , which is to be expected as the system is close to the mct glass transition .
we wish to mention here that collective motion was also observed in recent simulations on a 2-dimensional active hard - disk system @xcite . for larger self - propulsions ,
i.e. @xmath61 , the motion of the particles is less correlated as the system is well below @xmath24 .
hence , the dynamics becomes more fluid - like and the dynamic heterogeneities disappear in a system of active hard spheres with @xmath61 and @xmath56 .
we also study the effect of self - propulsion on the cage dynamics . to this end
, we measure the mean square displacement @xmath62 in a system of passive hard spheres at @xmath56 . the mean square displacement as shown in fig .
[ fig3]c shows a clear plateau at long time scales as the particles are kinetically trapped within the cages that are formed by their neighbours . to illustrate the cage effect , we present snapshots of a typical dynamical trajectory in fig .
[ fig3]f - g , where we highlight the particles that form the cage .
we clearly observe that the cage remains intact on at least a time scale of @xmath63 with @xmath64 the mean short - time diffusion coefficient . for comparison
, we also plot @xmath62 for a system of active hard spheres with @xmath65 , which shows diffusive behavior at short times , and subdiffusive behavior at intermediate time scales but without a clear cage - trapping plateau . at long time scales
, @xmath62 grows linearly with time and the particles become diffusive again with a long time diffusion coefficient @xmath66 .
the snapshots in fig .
[ fig3]d - e show that the cage has disappeared completely within a time scale of @xmath63 , which explains the absence of any cage effect in @xmath62 ( see supplementary movie 1 and 2 ) . finally , we investigated if a particle can break out of its cage by its own self - propulsion .
we performed a simulation of a single active particle with activity @xmath67 in a system of @xmath2 hard spheres at @xmath56 , and measured @xmath62 for both active and passive spheres .
we clearly find that the passive spheres are trapped within their cages as @xmath68 , whereas the single active hard sphere can jump out of its cage , although it remains a rare event .
if we increase the fraction of active spheres to @xmath69 , we find that the mobility of both the passive and active spheres increases significantly resulting in long - time diffusion coefficients @xmath66 and 0.0625 for the passive and active particles , respectively . we thus conclude that a motorized particle can occasionally break out of its cage by its own self - propelling force , but there is also a collective effect , because the mobility is even more enhanced upon increasing the fraction of active spheres . to shed light on the enhanced mobility with increased activity , we investigate the structure of the active hard - sphere system as a function of self - propulsion @xmath6 .
we calculate the static structure factor @xmath70 at @xmath56 and present our results in fig .
we clearly observe that the main peak of @xmath70 shifts to higher wave vector @xmath71 and becomes broader upon increasing the self - propulsion @xmath6 , which indicates that the average separation between neighboring particles becomes smaller , but the distribution of interparticle distances gets broader .
we thus conclude that the structure of an equilibrium glass of passive hard spheres is different from a non - equilibrium active hard - sphere system at the same packing fraction .
however , we also find that the difference in structure decreases upon increasing @xmath17 , and we speculate that the structure becomes similar in the limit of rcp as the system becomes fully jammed .
the observed structural changes can be explained by the fact that particles are pushed closer together due to the self - propulsion , thereby creating more `` free volume '' for other particles , which can then break out of their cages and the glass state can melt into a non - equilibrium fluid phase .
a similar melting of hard - sphere glasses was also observed in systems , where the addition of depletants instead of self - propulsions leads to a clustering of the particles and a destruction of the cages @xcite .
a similar picture emerged from a detailed simulation study on a 2-dimensional system of active hard disks , where it was found that active particles transiently stick to the neighbours found in the direction of their self - propulsion until the direction of the motion is changed due to brownian rotation @xcite . in this work
, we used a rotational diffusion coefficient @xmath72 according to the stokes - einstein relation . to study the effect of particle rotations on the dynamics of the system , we perform edbd simulations for a system of active hard spheres with @xmath59 and rotational diffusion coefficients @xmath73 varying from 0.3 to 300 @xmath74 .
we present the results in fig .
we observe clearly that the structural relaxation time increases monotonically with packing fraction , and that the dynamics is well - described by the vft and mct predictions for all values of @xmath73 that we studied .
note that @xmath75 corresponds to the stokes - einstein relation .
in addition , we find that upon increasing the rotational diffusion coefficient @xmath73 , the relaxation dynamics slows down , the critical mct and vft packing fractions @xmath24 and @xmath23 decrease , and the fragility index @xmath26 increases .
more specifically , we find that the relaxation dynamics of active hard spheres approaches that of passive brownian hard spheres in the limit of @xmath76 as the orientations of the self - propelling force at any time @xmath77 and @xmath78 ( @xmath79 ) become uncorrelated thereby recovering the white noise in brownian motion .
in conclusion , we performed extensive edbd simulations of systems of self - propelled motorized hard spheres , and investigated the dynamical behavior of systems that are far out of equilibrium .
we find that the relaxation dynamics can be sped up by orders of magnitude upon increasing the self - propulsion of the particles .
this result provides an important step forward towards a better understanding of dense packings as it allows us to investigate sphere packings with fluid - like dynamics at densities close to random close packing without crossing a glass transition .
for instance , the structural relaxation time of active hard spheres with a self - propulsion @xmath67 at a packing fraction as high as @xmath80 is similar to that of a passive hard - sphere system at @xmath81 .
moreover , we show that the non - equilibrium mct glass transition in a system of active hard spheres is pushed from @xmath82 volume fraction in the case of passive hard spheres to packing fractions close to rcp upon increasing the self - propulsion .
the non - equilibrium glass phase shows characteristics like dynamical arrest , cooperative motion , and a diverging structural relaxation time that are reminiscent of the equilibrium glass transition .
thus , our findings strongly support recent theoretical predictions on the existence of a non - equilibrium glass transition in active systems @xcite . consequently ,
as the packing fraction at which the system falls out of equilibrium depends on self - propulsion , and the resulting glass phase can only be compressed further to its corresponding jammed state at infinite pressure , we expect to find a range of jamming densities that depends on activity .
in addition , our structure factor data show that the structure of the equilibrium glass phase of passive hard spheres is different from that of active hard spheres , but we expect that the structure becomes similar in the limit of random close packing .
additionally , we show that the effect of a self - propulsion is more pronounced for slower rotational diffusion , while the dynamics of active hard spheres becomes brownian in the limit of an infinitely large rotational diffusion coefficient .
our study can also be extended to random packings of arbitrarily shaped particles , which will provide new insights in the concept of rcp as well as the glass transition of these systems .
it is also interesting to investigate if active particles can be used to unjam systems that are kinetically trapped in a glass phase in order to promote the formation of the thermodynamically stable phase .
we hope that our findings inspire future experimental investigations on glassy dynamics of dense suspensions of active particles or of colloidal glasses that only contain a small amount of active particles .
we wish to note here that most experimental studies so far have only been focussed on dilute instead of concentrated suspensions , and that most experimental systems correspond to much lower activities @xmath83 @xcite than studied here , but the light - activated colloids @xcite , the catalytic janus particles @xcite , and the particles with a artificial magnetic flagella @xcite are capable of producing self - propulsions as high as @xmath84 , 50 , 80 , respectively , which should have a profound effect on the relaxation dynamics in concentrated systems .
we perform event - driven brownian dynamics simulations @xcite of a system of self - propelled hard spheres . to avoid crystallization
we study a suspension of @xmath2 hard spheres with a gaussian size polydispersity of @xmath3 and a mean diameter of @xmath4 . even though the particles are driven and energy is continuously supplied to the system , the temperature @xmath36 is kept fixed as the solvent acts as a heat bath .
the overdamped motion of particle @xmath5 with position @xmath85 and orientation @xmath86 is given by @xmath87 where the potential energy @xmath88 is the sum of excluded - volume interactions between all hard spheres , and @xmath89 is the short - time self diffusion coefficient of particle @xmath5 .
a stochastic force with zero mean , @xmath90 , describes the collisions with the solvent molecules , and satisfies @xmath91 with @xmath92 the identity matrix .
in addition , the self - propulsion of particle @xmath5 is described by a constant force @xmath6 in the direction @xmath7 at time @xmath77 .
36ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1038/nphys2592 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) http://link.aps.org/doi/10.1103/physreve.76.041510 [ * * , ( ) ] @noop ( ) @noop * * , ( ) @noop * * , ( )
.the critical packing fraction @xmath23 and fit parameter @xmath26 as obtained from the vft fit ( @xmath93 $ ] ) of the structural relaxation time @xmath29 , and the critical packing fraction @xmath24 and fit parameter @xmath25 as extracted from the mct fit ( @xmath94 ) for active hard spheres with varying self - propulsions @xmath95 .
the errors are determined from a @xmath96 confidence interval for the fits . [ cols="^,^,^,^,^ " , ] | although the concept of random close packing with an almost universal packing fraction of @xmath0 for hard spheres was introduced more than half a century ago , there are still ongoing debates .
the main difficulty in searching the densest packing is that states with packing fractions beyond the glass transition at @xmath1 are inherently non - equilibrium systems , where the dynamics slows down with a structural relaxation time diverging with density ; hence , the random close packing is inaccessible .
here we perform simulations of self - propelled hard spheres , and we find that with increasing activity the relaxation dynamics can be sped up by orders of magnitude .
the glass transition shifts to higher packing fractions upon increasing the activity , allowing the study of sphere packings with fluid - like dynamics at packing fractions close to random close packing .
our study opens new possibilities of investigating dense packings and the glass transition in systems of hard particles .
random packings of hard spheres have been extensively studied as models for liquids , glasses , colloids , granular systems , and living cells , but they are also related to many mathematical problems , such as signal digitalization , error correcting codes , and optimization problems @xcite . nevertheless , there are ongoing debates on the characterization of amorphous packings of hard spheres , and the nature of random close packing ( rcp ) is still highly controversial @xcite .
one of the main issues is whether or not rcp is well - defined with a single value for the jamming density , also called the j - point , or by a finite range of jamming densities , a so - called j - line @xcite .
for instance , it has been argued that rcp is ill - defined as one can always obtain a higher packing fraction by introducing crystalline order in the system @xcite .
moreover , experiments and simulations show that a range of jamming densities can be obtained depending on the protocol that is used to generate random sphere packings @xcite .
the observed range of jamming densities has been explained by the rate at which accessible states disappear @xcite , and by the theoretical prediction that the jammed configurations can be regarded as the infinite - pressure limit of glassy states @xcite .
subsequent simulations indeed confirmed that a system falls out of equilibrium at a density where the relaxation time exceeds the compression rate , and that this glass phase can only be compressed further to its corresponding jammed state as structural relaxation is prohibited @xcite .
hence , a range of different jamming densities can thus be obtained as the glass transition depends on compression rate .
the dynamics of self - propelled ( active ) particles in dense suspensions has attracted an increasing amount of interest in order to describe self - organization phenomena like bird flocks , bacteria colonies , tissue repair , and cell cytoskeleton .
very recently , also artificial colloidal microswimmers have become available , which show promise for applications such as biosensing , drug delivery , etc .
@xcite .
examples of active self - propelled particles are colloids with magnetic beads that act as artificial flagella @xcite , catalytic janus particles @xcite , laser - heated metal - capped particles @xcite , light - activated catalytic colloidal surfers @xcite , platinum - loaded stomatocytes @xcite , or living bacteria @xcite .
in contrast to passive colloidal particles that only exhibit brownian motion due to thermal fluctuations , active self - propelled colloids are driven by the interplay between random fluctuations and active swimming .
while passive brownian particles are in thermodynamic equilibrium with the solvent , active particles are driven far from equilibrium as they convert incessantly energy into active motion .
recent experiments on crowded systems of active colloids and active cells show dynamical features like jamming and dynamical arrest that are very similar to those observed in glassy materials @xcite .
these experimental observations are in line with recent theoretical predictions @xcite that dynamic arrest can occur in systems that are far from equilibrium such as active and driven systems with characteristics that are very similar to those observed in glasses .
additionally , it was predicted theoretically that the non - equilibrium glass transition moves to lower temperature with increasing activity and to higher temperature with increased dissipation in spin glasses @xcite .
however , the temperature shift of the non - equilibrium glass transition could not be explained by replacing the temperature with an effective temperature @xcite . inspired by these results , here we investigate the dynamical behavior of dense suspensions of active particles and study whether or not a glass transition can occur in a system that is far out of equilibrium .
to exclude the effect of temperature , we investigate the effect of self - propulsion on dense packings of hard spheres that solely interact with excluded volume interactions .
remarkably , the glass transition shifts to higher packing fractions upon increasing the activity , which allows us to study for the first time random sphere packings with fluid - like dynamics at volume fractions close to rcp .
we also find that the effect of random self - propulsions decreases with larger rotational diffusion coefficients of the particles .
moreover , we confirm the existence of the non - equilibrium glass transition in active systems with a realistic particle model for the first time . |
it is well known that confining the dimensions of a physical system frequently leads to exciting new effects . for example confining electrons to a single dimension results in the breakdown of the fermi liquid model , and the observation of exotic transport behavior @xcite .
similarly , there is no long - range magnetic order in two dimensions @xcite . in @xmath7-electron materials , where magnetic and electronic degrees of freedom are typically strongly coupled , the effects of dimensionality are currently under debate @xcite .
notably , it has been proposed that dimension can be used to tune the type of observed quantum phase transitions , where either the electronic or magnetic degrees of freedom become quantum critical @xcite .
the entanglement of electronic and magnetic degrees of freedom in @xmath7-electron compounds is controlled via two magnetic interactions .
a large on - site coulomb repulsion tends to localize the @xmath7-electron wave function , producing magnetic moments that interact indirectly with each other via the ruderman - kittel - kasuya - yosida ( rkky ) interaction , which leads to a long - range antiferromagnetic order . on the other hand , below a characteristic temperature @xmath8 , the kondo interaction @xcite drives the local demagnetization of the @xmath7-electron state that is quenched by the spins of the surrounding conduction electrons .
this results in delocalization of the @xmath7-electrons into the conduction band and thus the formation of a paramagnetic heavy fermi liquid state @xcite .
the most interesting situation occurs when kondo and rkky interactions are of comparable size @xcite . at this point ,
the antiferromagnetic ordering temperature is suppressed to zero resulting in a quantum critical point ( qcp ) where the associated long - wavelength quantum critical magnetic fluctuations are believed to lead to the emergence of novel states of matter @xcite , with the most prominent one being unconventional superconductivity @xcite . apart from this prototypical qcp scenario @xcite
, so - called local quantum criticality has been proposed for the case that the kondo temperature is suppressed to zero at the qcp , resulting in local critical fluctuations associated with the fermi surface . in this scenario ,
the fermi surface is large on one side of the qcp because it includes the delocalized @xmath7-electrons , and is small on the other because the @xmath7-elctrons are localized and do not contribute to the fermi volume @xcite .
recent calculations suggest that the type of qcp may be controlled via the dimensionality of the system @xcite .
apart from influencing the type of qcp , lowering the dimensionality is also believed to result in an increase of the superconducting transition temperature @xcite .
this calls for new classes of @xmath7-electron materials in which the dimensionality can be systematically tuned .
we recently synthesized ce@xmath9al@xmath0si@xmath1 ( @xmath9 = rh , ir , pt ) compounds that are promising in this regard @xcite .
these materials crystallize in kcu@xmath0s@xmath10 structure type with tetragonal space group p4/mmm ( # 123 ) , and exhibit a sequential stacking of a baal@xmath0-type ce - containing layer separated by a aual@xmath1-type @xmath9-containing block along the @xmath6-axis ( fig .
[ fig1 ] ) .
notably , the structure can be described as @xmath11 = 1 member of ce@xmath12al@xmath13si@xmath1 , where @xmath11 is an integer .
increasing @xmath11 enlarges the thickness of the @xmath9al@xmath1 block and hence increases the distance between the ce planes containing the @xmath14 electrons without changing their crystal environment .
we note that the large family of @xmath15(aual@xmath1)@xmath16al@xmath1(au@xmath17si@xmath18 , where @xmath15 is a rare earth element compounds @xcite with @xmath11 = 0 - 3 has already been obtained as a disordered variant ( for @xmath19 = 0.5 ) of this structural motif ( @xmath19 = 0 ) .
al@xmath0si@xmath1 ( m = rh and ir ) . ]
bulk thermal , transport and magnetic properties reveal two low - temperature antiferromagnetic phases for cerhal@xmath0si@xmath1 ( @xmath2 = 14 k and @xmath3 = 9 k ) and ceiral@xmath0si@xmath1 ( @xmath2 = 16 k and @xmath3 = 14 k ) @xcite .
the effective magnetic moment of both cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 obtained in the paramagnetic region are close to ce@xmath20 hunds rule value ( 2.54 @xmath5 ) , indicating localized behavior of the ce @xmath14 moments at high temperature . however , the low temperature behavior point towards screening of these local moments by the kondo effect @xcite .
heat capacity measurements carried out down to 100 mk determined the sommerfeld coefficient to be @xmath21 = 200 mj / mol - k@xmath22 in cerhal@xmath0si@xmath1 and 50 mj / mol - k@xmath22 in ceiral@xmath0si@xmath1 , indicating moderately heavy fermion behavior . to make progress on the characterization of these materials ,
we have carried out neutron powder diffraction ( npd ) measurements on cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 to determine their magnetic structure .
single crystals of ce@xmath9al@xmath0si@xmath1 ( @xmath9 = rh , ir ) were grown from al / si flux @xcite , and ground into fine powders . for cerhal@xmath0si@xmath1 2.5 grams of powder
was loaded into a aluminum can with an inner diameter of 6 mm under a he gas atmosphere . in order to reduce the effects of neutron absorption due to
ir , the ceiral@xmath0si@xmath1 powder ( 3.5 g ) was loaded into an annular aluminum can , where the annulus was 1 mm thick .
the neutron powder diffraction experiments were carried out using the instrument hb2a at the high flux isotope reactor at oak ridge national laboratory .
neutrons of wavelength 2.413 , selected by a vertically focusing a ge(113 ) monochromator , were used for the experiment .
the calculated @xmath23 absorption length for this wavelength is 7.3 mm and 2.5 mm for cerhal@xmath24si@xmath25 and ceiral@xmath24si@xmath25 , respectively , matching the diameters of the selected aluminum cans . for both compounds ,
powder diffraction patters were collected for 6 hours for each of the two magnetic phases as well as slightly above the higher temperature magnetic phase .
in addition , the temperature variation of the intensity of the strongest magnetic peak was also measured to obtain the temperature dependence of the magnetic order parameter .
at each temperature , the peak intensity was measured for three and six minutes for cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 , respectively .
finally , several npd patterns were collected while cooling from room temperature ( 100 k ) to the base temperature of 3.2 k in the case of cerhal@xmath0si@xmath1 ( ceiral@xmath0si@xmath1 ) to check for structural distortions .
rietveld refinement @xcite of both the crystal structure and magnetic structure from the npd data was carried out using the fullprof software @xcite .
the magnetic structures allowed by symmetry were obtained via a representational analysis using the software sarah @xcite .
si@xmath1 and b ) ceiral@xmath0si@xmath1 measured above the magnetic transition temperature .
the al peaks from the sample holder are indicated by an asterisk ( * ) , which were modeled by constant profile matching ( le bail fit ) . ]
figs . [ fig2](a ) and ( b ) show the rietveld refinement of the neutron powder diffraction patterns of cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 collected in the paramagnetic state at @xmath26 = 20 and 30 k , respectively .
the aluminum peaks from the sample holder are marked by asterisks .
the results of the refinements are presented in table [ table1 ] . in each case , the crystal structure deduced from single crystal x - ray diffraction @xcite was used as a starting model . for both samples , no new bragg peaks appeared during cooling , and within the resolution of the collected data , no indication of a structural distortion was found .
although the samples showed no impurity peaks in the room temperature x - ray patterns , small additional peaks were observed in ceiral@xmath0si@xmath1 at all temperatures , suggesting that they can be attributed to a small unknown nonmagnetic impurity phase .
we note that the refinement is better in cerhal@xmath0si@xmath1 than in ceiral@xmath0si@xmath1 , most likely because the annular sample can used for ceiral@xmath0si@xmath1 may lead to a slight distortion of the peak shapes and intensities .
si@xmath1 measured at 20 k ( pink curve ) and at 3.2 k ( black curve ) .
the green solid line shows the difference curve obtained from these two profiles .
b ) rietveld refinement of the magnetic peaks at _ t _ = 3.2 k obtained from the difference pattern @xmath27 = @xmath28 - @xmath29 .
the lower ( orange ) plot is the difference pattern @xmath27 = @xmath30 - @xmath29 shifted by 12 below zero .
the features marked by asterisks are subtraction artifacts due to the temperature dependence of the debye - waller factor . ]
.crystallographic parameters and agreement factors from rietveld refinements in space group p4/mmm with ce atoms at ( 0 , 0 , 1/2 ) , m = rh , ir at ( 0 , 0 , 0 ) , al atoms at ( 0 , 0 , z@xmath31 ) and si atoms at ( 1/2 , 1/2 , z@xmath32 ) . [ cols="<,<,<",options="header " , ] the four magnetic bragg peaks observed in cemal@xmath0si@xmath1 ( m = rh , ir ) for temperatures @xmath33 may be successfully indexed with a magnetic propagation vector @xmath4 = ( 0 , 0 , 1/2 ) , indicating a doubling of the magnetic unit cell along the crystallographic @xmath6-axis compared to the structural unit cell . using representational analysis based on the identified propagation vector and the space group p4/mmm yields two magnetic irreducible representations ( ir ) for the ce site that are summarized in table [ table2 ] . here ,
the @xmath34 and @xmath35 irreducible representations describe solutions with magnetic moments parallel and perpendicular to the @xmath6-axis , respectively .
a rietveld refinement was carried out on the magnetic contribution to the diffraction patterns recorded at @xmath26 = 3.2 k for cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 , as shown in figs . [ fig3](b ) and [ fig4](b ) ( red lines ) , respectively . in agreement with magnetic susceptibility measurements that indicate easy - axis anisotropy along the @xmath6-axis @xcite ,
the data is best described by the @xmath34 ir , in which ferromagnetic sheets of ce magnetic moments are stacked antiferromagnetically along @xmath6 .
the corresponding magnetic structure is illustrated in fig .
[ fig5 ] .
the rietveld fit yields ordered moments of @xmath36 = 1.14(2 ) and 1.41(3 ) @xmath5/ce for cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 , respectively .
the temperature dependence of the staggered magnetization for both compounds was obtained on the strongest magnetic peak and is shown in fig .
the onset of long - range antiferromagnetic order occurs below @xmath3 , as signaled by the non - zero magnetic moment .
si@xmath1 at _ t _ = 3.2 k determined from the neutron powder diffraction experiment .
the solid ( yellow ) balls are the ce atoms and the arrow represents the direction of the magnetic moment . ]
si@xmath1 and b ) ceiral@xmath0si@xmath1 .
magnetic moments at 3.2 k were obtained from rietvield refinement of the magnetic bragg peaks as discussed in the text .
the magnetic moments at all other temperatures were obtained by scaling the moment determined at 3.2 k with the square root of the intensity of the strongest peak measured at the respective temperatures . ] in summary , the ground state magnetic structure in both cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 at 3.2 k was determined to be a commensurate collinear antiferromagnetic structure with propagation vector @xmath4 = ( 0 , 0 , 1/2 ) via neutron powder diffraction .
although two magnetic transitions were clearly found in bulk properties measurements , no magnetic bragg peaks were observed between @xmath2 and @xmath3 .
possible reasons are that the moments in the region @xmath37 are too small to be observed within neutron powder diffraction , or that the magnetic structure of the high temperature magnetic phase is of an incommensurate nature .
we note that both cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 contain elements with relatively high neutron absorption , which makes the observation of small magnetic moments difficult .
methods that are typically more sensitive to small ordered magnetic moments such as single crystal neutron diffraction or nuclear magnetic resonance measurements will be required to determine the nature of the magnetic order in between @xmath2 and @xmath3 .
the size of the magnetic moments of @xmath38= 1.14(2 ) and 1.41(3 ) @xmath5/ce for cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 , respectively , obtained here are consistent with the saturated values of magnetic moments determined via magnetization measurements in the field polarized state ( 0.95 and 1.14 @xmath5/ce for cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 , respectively ) @xcite .
moreover , the magnitude of the ordered magnetic moments are much smaller than than the moment size of a free ce@xmath20 ion given by @xmath39 = 2.14 @xmath5 .
typically , a reduction of the ordered moment in rare earth systems may be due to crystal field effects and/or to kondo screening .
_ @xcite suggested that crystal field effects alone can not describe the reduced size of the magnetic moment and attributed the reduction of the moment to the kondo screening of the @xmath14 moments .
this is supported by heat capacity measurements @xcite , which indicate that the sommerfeld coefficient is large in cerhal@xmath0si@xmath1 ( @xmath40 mj / molk - k@xmath41 ) , and moderately enhanced in ceiral@xmath0si@xmath1 ( @xmath42 mj / molk - k@xmath41 ) , implying that kondo screening is indeed an important effect in these compounds .
we note that ultimately a determination of the crystal electric field parameters by neutron or x - ray scattering will be useful to elucidate the relative roles of kondo and crystal fields effects on the size of the ordered moment . assuming that the kondo effect is indeed the main contributor in reducing the size of the magnetic moment
, both cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 appear to be promising candidates to investigate the interplay of electronic and magnetic degrees of freedom near a qcp while tuning the dimensionality of the magnetic interactions . because of its lower magnetic ordering temperature and larger sommerfeld coefficient ( i.e. larger kondo interaction ) , cerhal@xmath0si@xmath1 is likely to be nearer to a putative antiferromagnetic qcp
. therefore measurements of its magnetic properties as function of chemical substitution or external pressure are highly desirable in order to test whether these novel compounds can be tuned towards a qcp .
work at los alamos national laboratory was performed under the auspices of the us department of energy , office of basic energy sciences , division of materials sciences and engineering .
research conducted at ornl s high flux isotope reactor was sponsored by the scientific user facilities division , office of basic energy sciences , us doe . | we have investigated the magnetic ground state of the antiferromagnetic kondo - lattice compounds cemal@xmath0si@xmath1 ( m = rh , ir ) using neutron powder diffraction . although both of these compounds show two magnetic transitions @xmath2 and @xmath3 in the bulk properties measurements , evidence for magnetic long - range order was only found below the lower transition @xmath3 .
analysis of the diffraction profiles reveals a commensurate antiferromagnetic structure with a propagation vector @xmath4= ( 0 , 0 , 1/2 ) .
the magnetic moment in the ordered state of cerhal@xmath0si@xmath1 and ceiral@xmath0si@xmath1 were determined to be 1.14(2 ) and 1.41(3 ) @xmath5/ce , respectively , and are parallel to the crystallographic @xmath6-axis in agreement with magnetic susceptibility measurements . |
since long , the various properties of excited nuclear matter occurring in heavy collisions have been studied within statistical descriptions @xcite . here
the underlying principle is the stringent assumption of full thermal equilibrium in the stage of the reaction to which the description is applied . in the last 10 years
thermal model ansatzes of hadronic resonance gases via a grand canonical description have become very popular again because detailed experimental results for the yields of individual hadronic species became available at various bombarding energies .
indeed , it was found that the statistical description can be applied very successfully for ultrarelativistic energies , indicating that the individual hadronic particles seem to evolve to a point of nearly perfect chemical equilibrium @xcite ( for a comprehensive review the reader is referred to @xcite ) .
however , for low collision energies ( e.g. in the sis , ags and fair range ) especially the strange particles do become rare .
thus , as has been well known for a long time , a canonical instead of a grand canonical description of the strange particles has to be applied .
the canonical approach results in a suppression in the yields for rare particles in comparison to the grand canonical description when treating the conservation of the corresponding u(1)-charge exactly @xcite .
indeed , thermal models including the canonical suppression have been applied rather successfully for the description of the few measured hadronic yields in heavy ion collisions at sis energies @xcite .
a dynamical interpretation of the canonical suppression has recently also been offered by the formulation and solution of kinetic master - equations @xcite .
it is the purpose of this investigation to show that such an occurrence of canonical suppression for rare particles is already warranted by present day transport models .
a priori this is not a trivial statement .
one could argue that the transport models are based on solving a set of coupled boltzmann equations which originate from a grand canonical treatment . however , this is not the case as the realisations do conserve energy exactly and also the individual quantum charges like baryon number and ( net ) strangeness are conserved during the propagation and scattering processes as well . in the following section we briefly review the canonical suppression of rare particles associated with the conservation of an u(1)-charge and we also summarize its dynamical formulation via a master equation as given in @xcite . in section iii
we describe our simulation setup .
a pion gas is simulated by incorporating inelastic and volume - limited collisions @xmath0 for the production of kaon pairs .
we employ two completely different numerical realisations for the treatment of collisions .
the first model under investigation is urqmd @xcite , where two - particle collisions are realized via a standard geometrical interpretation . for previous studies of the thermodynamic properties of hadronic matter within the urqmd model ,
the reader is referred to @xcite . in the second model
the collisions are treated by transition rates in small spatial subcells .
the latter algorithm has most recently been successfully introduced in a covariant parton cascade to describe inelastic multiparticle bremsstrahlung processes of type @xmath1 @xcite . in section
iv we present the results of our analysis .
section v then provides a summary and a conclusion .
we will argue that a ( canonical ) chemical equilibrium of kaons can by far not be achieved at intermediate energies for relativistic heavy ion collisions , at least if known cross sections for the production of kaons are assumed .
the statistical description of systems incorporating the exact conservation of quantum numbers has been established for many years and there exist several approaches of varying complexity and generality @xcite . here
we give a brief summary of the results adapted to our needs .
the conserved u(1)-charge to be considered is strangeness whose net value is taken to be zero throughout our calculations and simulations .
explicitly we consider only inelastic reactions of the type @xmath2 , where the kaons and anti - kaons bear strangeness @xmath3 and @xmath4 respectively . for a large number of kaons it is sufficient to treat strangeness conservation on the average @xmath5 ( with @xmath6 , @xmath7 being the number of kaons and anti - kaons , respectively ) .
introducing a chemical potential @xmath8 , or fugacity @xmath9 , to control the net strangeness content , one is led to the following grand canonical partition function @xmath10 where @xmath11 denotes the relativistic one - particle partition function for non - interacting particles of type @xmath12 ( pions or kaons in this case ) , @xmath13 is the volume , @xmath14 is the temperature and @xmath15 is the particle s mass .
@xmath16 denotes a modifed bessel function . @xmath17 in order to consider a more general case , the one - body partition functions in ( [ gcz ] ) would have to be replaced by sums over one - body partition functions of all hadronic particles carrying the corresponding integer in strangeness of @xmath18 , @xmath19 , @xmath20 and @xmath21 .
a small number of kaons ( roughly spoken when the average number becomes of order one ) demands the conservation of strangeness to be treated exactly , i.e. @xmath22 for each state contributing to the partition sum .
this constraint leads to a reduction of the available phase - space for the production process and ultimately one obtains the canonical partition function describing the system via ( @xmath23 and @xmath24 denote bessel functions ) @xmath25 with @xmath26 from ( [ gcz ] ) and ( [ cz ] ) one can calculate the density @xmath27 of particles with strangeness @xmath3 , the kaons , via the relation : @xmath28 and considering the canonical case one finds @xmath29 using that in the grand canonical picture @xmath5 requires @xmath30 . comparing ( [ gcdens ] ) and ( [ cdens ] ) one defines the canonical suppression factor @xmath31 @xmath32 which contains all relevant information on differences in the equilibrium particle density between the grand canonical and canonical description . an alternative way of understanding this suppression is to consider kinetic master equations @xcite by looking at a single process @xmath33 .
@xmath34 denotes the probability to find @xmath35 particles @xmath36 at a time @xmath37 . in our case
we have @xmath38 and thus @xmath39 holds exactly , whereas particles @xmath40 and @xmath41 ( the ` pions ' ) are assumed to be uncorrelated and abundant .
furthermore , the probabilities for a single gain or loss process per unit time and volume are denoted by @xmath42 and @xmath43 respectively , where @xmath44 and @xmath45 are the momentum - averaged cross sections for the gain and loss processes . with that
, a master equation can be formulated @xcite @xmath46 a general solution of this equation is possible @xcite , but for our purpose it is sufficient to look at the kinetic equation for the time evolution of the average number @xmath47 .
it can be obtained when multiplying ( [ mastereq ] ) by @xmath35 and summing over it : @xmath48 this equation can be easily treated in the two limiting cases @xmath49 and @xmath50 . for abundant production of @xmath51 pairs , i.e. when @xmath49 , the relation @xmath52 holds . on the other hand , for very rare production ,
i.e. when @xmath50 , one has the relation @xmath53 . assuming a thermal momentum distribution with @xmath54 and looking at stationary solutions for large times ( the solutions describing an equilibrated system ) one ends up with ( @xmath55 ) @xmath56 for @xmath49 .
the opposite case @xmath50 leads to @xmath57 identifying particle - type @xmath36 with kaons , it is clear that ( [ kinsolgc ] ) equals the grand canonical result ( [ gcdens ] ) .
( [ kinsolc ] ) is just the leading term in an expansion of the canonical result ( [ cdens ] ) .
hence , it is verified that an abundant production of particles leads to a grand canonical description , whereas rare particle production requires a canonical description .
let us have a closer look at the canonical suppression factor ( [ eta ] ) using the asymptotic behaviours @xmath58 here , one sees that in the grand canonical limit the kaon density @xmath59 is independent of the reaction volume , whereas in the canonical regime with the number of kaons @xmath60 , it scales linearly with the volume as @xmath61 .
[ fig_denstheory ] illustrates this behaviour . .
g.c . denotes the grand canonical calculation , c. labels the canonical result . ]
the main objective of our study is to demonstrate that current ( large scale ) transport models are able to reproduce the effect of canonical suppression .
this is not a trivial investigation .
one might argue that the transport algorithms are based on strictly solving the boltzmann equation which in turn originates from a grand canonical treatment . in the stationary limit of the ( coupled ) boltzmann equations the equilibrium phase space distributions , where the various collision terms for each particle population vanish , obey maxwell - boltzmann statistics ( or bose - einstein or fermi - dirac statistics when the pauli factors are included ) , so that the individual densities are that of a grand canonical ensemble .
the reason for this is that for describing an annihilation within a boltzmann description the two ( rare ) particles are assumed to be completely uncorrelated .
this is however not the case in the description via the master equation ( [ mastereq ] ) . here
the occurrence of a pair of rare particles is explicitly taken care of . due to the nature of pair production the existence of a kaon
requires the existence of an anti - kaon .
this leads to an effective increase in the annihilation process compared to the boltzmann stosszahlansatz . inspecting the loss term in ( [ kineq ] ) , the approximation leading to a boltzmann process
is given by @xmath62 which then leads to the grand canonical description .
staying within the description of the master equation , the possibility for annihilation is much higher .
when there exists a kaon , there has to be an anti - kaon with which an annihilation process is possible .
@xmath63 for @xmath64 , hence the effective annihilation rate incorporated correctly in the master equation is much larger than for a boltzmann description , leading to the canonical suppression . on the other hand the various numerical realisations of the underlying transport equations do conserve energy and the individual quantum charges ( like baryon number and net strangeness ) exactly - as within a microcanonical treatment .
the argument of an ` enhanced ' annihilation should therefore also apply within these realisations . in order to address this question ,
we will concentrate on the volume dependencies in the grand canonical and canonical regime respectively , as depicted in fig .
[ fig_denstheory ] .
the simulation setup consists of a large box of @xmath65 side length holding a relativistic gas of pions .
the pion gas provides a heat bath for a much smaller reaction volume of variable size centered within the large box . inside this smaller and likewise box - shaped reaction volume processes @xmath66 are allowed , covering all possible isospin states of pions and kaons .
the kaons are reflected by the walls of the small reaction volume and are thus bound to it , whereas these walls are permeable for the pions .
an illustration of this special spatial setting is given in figure [ fig_illustration ] .
after equilibration , the kaon density within the reaction volume should be governed by eq .
( [ cdens ] ) which holds as a reference for the transport model results .
the idea therefore is to simulate different sizes of the inner reaction volume and to extract the number of kaons @xmath67 for each simulation run by averaging over many timesteps in order to minimise the statistical fluctuations .
these timesteps are sufficiently separated to avoid correlations .
furthermore the data are taken after the equilibration of the system .
we implemented this scenario using two different types of transport descriptions - the microscopic transport model urqmd @xcite and a realisation of a stochastic transport model borrowed from a recently developed parton cascade @xcite .
the former model makes use of a geometrical interpretation of cross sections in order to solve the transport equations , whereas the latter relies on the explicit calculation of transition probabilities . for standard applications ,
the urqmd model provides full space time dynamics for hadrons and strings .
it is a non - equilibrium model based on the covariant propagation of hadrons and strings .
all cross sections are fitted to available data or calculated by the principle of detailed balance . for our studies the code
is modified such that only reactions @xmath2 , together with elastic collisions among the pions and kaons , remain possible .
the @xmath68 inelastic reactions are assigned a constant cross section of either @xmath69 , @xmath70 or @xmath71 , whereas the backward reactions are calculated via the principle of detailed balance .
the simulation of @xmath72 processes within the stochastic method is based on the calculation of a collision probability for each possible particle pair per unit volume @xmath73 and unit time @xmath74 via @xcite @xmath75 here , @xmath76 denotes the relative velocity and @xmath77 is the cross section for the considered @xmath72 process .
similar to the urqmd setup , the cross sections are set to be constant in one direction and calculated via detailed balance for the reverse reaction .
the so obtained probability is then compared with a random number between @xmath18 and @xmath78 to decide whether the collision does take place or not .
the implementation of the stochastic model is therefore closely related to the formulation of the master equation ( [ mastereq ] ) and its solution discussed in section [ sectheory ] .
we have to stress one important point : the stochastic method is , in principle , flexible to introduce test particles , that is to ` subdivide ' each particle into a number n of testparticles @xcite .
however , for the following investigation it is crucial that the produced kaons are _ not _ subdivided into further testparticles .
we will discuss this further below .
the initial conditions in both schemes are chosen such that the pion gas acquires a temperature of @xmath79 .
the appropriate number of pions and the total energy of the system are calculated via the use of a grand canonical partition function ( [ gcz ] ) for pions alone , as kaons are absent in the initial state .
the total energy evaluates to @xmath80 a heat bath volume of @xmath81 , as used in our simulations , then corresponds to a population of @xmath82 pions bearing a total energy of @xmath83 .
initially each pion is assigned the same fraction of the total energy , giving one half of the particles momenta in the positive @xmath84-direction , while the remaining particles start out bearing momenta in the negative @xmath84-direction .
the spatial distribution is random .
a necessary verification of the simulations reliability is to check for kinetic equilibration .
[ fig_spectrum ] demonstrates that the energy distribution of the pions does become exponential over six orders of magnitude with the desired temperature of @xmath79 in both schemes .
[ hbt ] figures [ fig_denssimcascade ] and [ fig_denssimurqmd ] now display the actual results in terms of the kaon density as a function of the ( small ) reaction volume .
the minor fluctuations in the results indicate the small statistical errors due to the finite number of averages done .
it is clear that the canonical suppression is reproduced by both transport schemes .
the kaon yield is suppressed for small reaction volumes with respect to the grand canonical limit .
this verification states the main result of the present study .
[ hbt ] ) for a temperature of @xmath85 and corrected by taking small deviations in temperature and fugacity into account.,title="fig : " ] [ hbt ] ( triangles ) , @xmath86 ( circles ) and @xmath71 ( diamonds ) have been assigned to the forward direction of the inelastic reactions @xmath0 .
the solid and dashed lines again provide the theoretically expected behaviour as seen in fig .
[ fig_denssimcascade].,title="fig : " ] in general , certain deviations from the theoretical values are to be expected due to the choice of initial conditions .
pion number and energy content are fixed such that , without any kaons present , the system is tuned to the reference temperature of @xmath79 and fugacity @xmath87 .
the production of the heavier kaons then leads to small changes in these parameters , as the total energy in the system is conserved as well as the total number of particles , because only particle number conserving collision processes are included in the present study .
this slightly affects the equilibrium number of pions and thus also the kaons . for the case of the stochastic cascade we investigated the relative deviations of temperature @xmath88 and effective mesonic fugacity @xmath89 more closely .
it is found that both deviations indeed grow with the number of kaons produced and thus with the size of the reaction volume .
the observed deviations are on the order of two percent at most . taking these deviations into account
, the theoretical results can be adapted to the actual conditions present in the system at a given size of the reaction volume by rescaling with the effective fugacity factor .
the actual theoretical reference curves in figs [ fig_denssimcascade ] and [ fig_denssimurqmd ] are corrected that way .
the stronger deviations for the microscopical urqmd model can not be solely explained by the changes in temperature and fugacity .
in fact the crucial effect when using the geometrical concept of incorporating binary collisions is a decrease in the collision rate , when the expected mean free path @xmath90 for the particular reaction gets in the order of the interaction length @xmath91 as pointed out in @xcite .
this is due to the difference in the collision times of the involved particles viewed from the computational frame . during that interval
the particle having the larger collision time must not collide again to ensure causality .
thus , the collision rate is decreased compared to the one given by the collision integral . as different densities
are involved , one can not expect forward and reverse reactions to be affected in the same way and the changes in reaction rates lead to a shifted stationary ` equilibrium ' value of the kaon density .
this behaviour can be clearly seen in fig .
[ fig_denssimurqmd ] , where the kaon density is depicted for different cross sections of the forward reaction @xmath68 .
the smaller the chosen cross section and thus the larger the mean free path , the more accurately the theoretical result is reproduced . on the contrary , for typical real mesonic cross section
the numerical shortcoming is in the range of 10 to 15 percent .
too few kaons relative to the pions are simulated in comparison to the theoretical limit either in the canonical or grand canonical regime .
such a discrepancy in kaon number relative to the pion number with respect to experimental results has been reported in several urqmd calculations @xcite .
it might well be that a better implementation of the collision criteria will enhance the kaon relative to the pion yield .
we leave this for a detailed future investigation . another instructive way to analyse the results is to look at the suppression factor @xmath92 ( [ eta ] ) as a function of the average number of kaons @xmath93 present in the system .
this relation is independent of the temperature which can be easily shown : since the suppression factor is a function of @xmath94 , @xmath95 , the one - body partion function can in principle be expressed in terms of @xmath92 , @xmath96 which then leads to @xmath97 [ hbt ] as a function of the average kaon number.,title="fig : " ] fig .
[ fig_nvseta ] states the summary of the investigation and depicts the above functional relation together with the simulation results . here
the temperature and fugacity deviations have been taken into account for the cascade results .
for the calculations employing the urqmd model the suppression factor has been calculated by taking the ratio of the actual kaon density @xmath98 divided by the density @xmath99 in the limit of very large volumes . within this prescription the sensitivity on the cross section vanishes almost completely .
apart from the demonstration of the excellent agreement between theory and simulation , the plot illustrates the two limiting cases considered when solving the kinetic equation ( [ kineq ] ) .
small particle numbers @xmath100 are associated with the canonical regime , whereas for abundant particle production @xmath101 the grand canonical description is fully valid .
we have demonstrated that two current ( large scale ) transport models are able to reproduce the effect of canonical suppression . as a particular example
we have chosen the production of kaons and anti - kaons via pions .
the overall kaon yield is suppressed with respect to the grand canonical limit for small reaction volumes and thus small numbers of produced kaons .
the transport algorithms are based on solving the boltzmann equation .
still they do conserve energy and net strangeness exactly .
the canonical suppression for kaon numbers @xmath67 considerably smaller than one , as pointed out in @xcite , originates dynamically from an enhancement of the annihilation process by @xmath102 as compared to standard , grand canonical formulation of the boltzmann equation .
the reason is that any kaon in the system requires the existence of a corresponding anti - kaon due to strangeness conservation .
thus the probability of finding a particle anti - particle pair turns into a highly correlated conditional probability compared to that obtained via the uncorrelated boltzmann stosszahlansatz .
the so enhanced annihilation probability then leads to the canonical suppression in the kaon yields . as the calculations show , it is manifest that the occurrence of canonical suppression is dynamically reproduced by the two presented schemes for solving the kinetic transport equations .
a small discrepancy , i.e. a slightly lower kaon yield on the order of 10 to 15 percent , is present when employing typical cross sections .
this discrepancy can be traced back to the numerical realisation of the occurrence of collisions via the standard geometrical description . as already emphasized , detailed investigations within the urqmd model
are in order to clarify how important such effects are for the understanding of kaon yields in relativistic heavy ion collisions , as typically too few kaons are produced relative to the pions , when compared to experimental results @xcite . addressing especially the dynamical generation of canonical suppression
, one has to worry about even more significant violations when other numerical transport schemes are invoked : one popular scheme is the so called test - particle method , i.e. each real particle is subdivided into @xmath103 test - particles in order to obtain smoother distributions or to suppress other numerical artefacts ( see e.g. @xcite ) .
when applying this idea of particle subdivision to the stochastic method , it will rescale the volumes v of figure [ fig_denssimcascade ] by a factor @xmath104 , so that the grand canonical limit is achieved for much smaller volumes . this is , of course , unphysical .
thus , whenever the canonical suppression is of relevance for the understanding of particular yields , the test - particle method has to be avoided .
another oftenly applied method is the so called perturbative method with ` virtual ' particles ( for a discussion see e.g. @xcite ) .
here it is assumed that the exotic particles being produced are so rare that their behaviour does not alter the overall dynamics . if the backreaction of kaon production , i.e. the annihilation of a kaon and its antiparticle , is not considered correctly , then detailed balance is violated and the system can never achieve full and correct chemical equilibrium .
when simulating real heavy ion collisions to address the production of kaons for lower or moderate relativistic energies ( for a very recent work see @xcite ) , one has indeed to worry whether these backreactions are important in the sense that the rates to achieve chemical equilibrium are smaller or comparable to the overall lifetime of the fireball .
various applications of thermal models for such low energies have been put forward in several works over the last years for describing hadronic yields in heavy ion collisions at sis @xcite .
as only very few yields are available , the soundness of such an analysis is not evident . in principle
, such an analysis rests on the assumption that chemical equilibrium among the individual hadronic particles is temporarily achieved , which implies that the corresponding time scales are sufficiently short compared to the lifetime of the fireball . from transport theory residing on binary elastic and inelastic hadronic collisions
it has been known for a long time that for intermediate to moderate energies the kaons are predominantly produced before the initial motion is substantially degraded and when the system is still far from any ( quasi-)equilibrium stage ( for a discussion and review see e.g. @xcite ) . putting it differently , when the momenta of the nucleons are sufficiently degraded and the system has to some extent thermalized , the timescale for production of strange particles via the considered inelastic kinetic reactions becomes extremely large .
for a particular calculation in a static environment at somewhat larger energies it was found that the equilibration of the kaons is exceeding the lifetime of a potential fireball by at least two orders of magnitude @xcite . in this analysis
the comparison was made , however , to a grand canonical estimate . on the other hand ,
as we will argue now , a chemical equilibrium situation of kaons with canonical suppression included can not be achieved even temporarily when assuming known cross sections . within a grand canonical environment
the rate for chemical equilibration of a particle c for a reaction of type @xmath105 is given by @xmath106 , where @xmath107 is the particle density of species @xmath108 . with @xmath109
if there is only one kaon , then there also has to exist one anti - kaon in the particular realisation of the system .
hence , for such a situation of a canonical realisation , where the reactions @xmath110 will drive the kaons to chemical equilibrium , the subsequent equilibration rate is obtained by substituting for @xmath107 the effective anti - kaon density @xmath111 .
the rate thus reads @xmath112 in an urqmd simulation the typical averaged cross section rather close to threshold is given by @xmath113 , so that @xmath114 .
any typical reaction volume at intermediate time scales in a relativistic heavy ion reaction is in the order of 100 @xmath115 .
hence , kaons can not be expected to come or stay close to chemical equilibrium .
the authors thank s. leupold and u. mosel for discussions , which have stimulated this work . this work is supported by gsi , dfg and bmbf .
the computational resources have been provided by the center for scientific computing in frankfurt .
braun - munzinger , j. stachel , j. p. wessels and n. xu , phys .
b * 365 * ( 1996 ) .
p. braun - munzinger , d. magestro , k. redlich and j. stachel , phys .
b * 518 * , 41 ( 2001 ) p. braun - munzinger , k. redlich and j. stachel , arxiv : nucl - th/0304013 .
l. v. bravina _ et al .
_ , phys .
c * 60 * ( 1999 ) 024904 [ arxiv : hep - ph/9906548 ] .
l. v. bravina _ et al .
_ , phys . rev .
c * 63 * ( 2001 ) 064902 [ arxiv : hep - ph/0010172 ] .
l. v. bravina _ et al . _ ,
c * 62 * ( 2000 ) 064906 [ arxiv : nucl - th/0011011 ] . | we demonstrate the occurrence of canonical suppression associated with the conservation of an u(1)-charge in current transport models . for this study
a pion gas is simulated within two different transport approaches by incorporating inelastic and volume - limited collisions @xmath0 for the production of kaon pairs .
both descriptions can dynamically account for the suppression in the yields of rare strange particles in a limited box , being in full accordance with a canonical statistical description . |
the topological notion of covering dimension was extended to the noncommutative context by winter and zacharias in @xcite . their _ nuclear dimension _
has contributed to the most important and broadly applicable advances in the theory of nuclear @xmath0-algebras in the last 40 years .
for example , the toms winter conjecture asserts that finite nuclear dimension is often equivalent to structural properties analogous to those exploited by connes in his proof of uniqueness of the injective ii@xmath4-factor ( @xcite ) . in a remarkable breakthrough
, this audacious conjecture was confirmed in the unique - trace case in @xcite , and has now been confirmed for much broader classes ( see @xcite ) . as another stunning example , in 2015 finite nuclear dimension led to the completion of elliott s classification program for the cases of most interest ( cf .
@xcite , @xcite ) . in short ,
nuclear dimension has revolutionized the field .
it is thus important to know which examples have finite nuclear dimension . in the influential paper
@xcite it was shown that all kirchberg algebras have nuclear dimension at most three .
( in fact , they have dimension one ; see @xcite and @xcite . ) in @xcite szabo proved that if @xmath5 acts freely on a compact metric space @xmath6 of finite covering dimension , then the associated crossed product @xmath7 has finite nuclear dimension .
his work was inspired by and depended upon @xcite , where the classical measure - theoretic rokhlin property was exported to the realm of topology .
this so - called _ rokhlin dimension _ makes sense in the noncommutative context too , so we can study it for @xmath0-dynamical systems .
there is mounting evidence that rokhlin dimension is for @xmath0-dynamical systems what nuclear dimension is for @xmath0-algebras : ubiquitous and fundamental .
for example , it is intimately connected with nuclear dimension via the crossed product construction .
one of the main results of @xcite was that @xmath8 has finite nuclear dimension whenever @xmath1 does and the automorphism @xmath9 has finite rokhlin dimension . since @xmath8 can be realized as a cuntz pimsner algebra , it is natural to seek an extension of rokhlin dimension to the context of @xmath0-correspondences , then ask whether the associated toeplitz pimnser and/or cuntz
pimsner algebras have finite nuclear dimension ? that is the subject of this paper .
indeed , we generalize rokhlin dimension to @xmath0-correspondences in definition [ rokhlin - dimension - def ] , then prove the following . [
main - theorem ] suppose that @xmath2 is a countably generated @xmath0-correspondence over a separable unital @xmath0-algebra @xmath1 , satisfying a technical quasidiagonal - like condition ( see theorem [ main_thm ] )
. then @xmath10 the quasidiagonal - like condition is probably unnecessary , and would be if a certain algebra could be shown directly to have finite nuclear dimension ( cf .
lemma [ quasi_hypothesis ] ) . in any case
, it is satisfied in many examples , including all finitely generated projective correspondences ( see example [ qd - examples ] ) . as an application
, we provide new examples of classifiable @xmath0-algebras in the sense of elliott s program .
thanks to @xcite and @xcite , this amounts to verifying simplicity , finite nuclear dimension and the universal coefficient theorem ( uct ) of @xcite .
( corollary [ application2 ] ) [ application ] assume @xmath1 is simple , unital , satisfies the uct and has finite nuclear dimension . for every finitely generated projective @xmath2 with finite rokhlin dimension , the associated cuntz
pimsner algebra @xmath11 is also simple , unital , satisfies the uct and has finite nuclear dimension .
note that we ve substantially generalized the @xmath0-dynamical system case from ( * ? ? ?
* theorem 4.1 ) because @xmath8 is the cuntz pimsner algebra over a singly generated projective correspondence .
also , this corollary holds whenever theorem [ main - theorem ] does , so it s likely true for arbitrary correspondences of finite rokhlin dimension ( and definitely true when the quasidiagonal - like hypothesis is satisfied ) .
the proof of theorem [ main - theorem ] follows @xcite very closely , at least in spirit .
the main technical innovation is finding suitable replacements for the outgoing and incoming maps used in the proof of theorem 4.1 in @xcite .
but we also have to worry about the nuclear dimension of the range of the outgoing maps , an easy task in the crossed - product case .
the majority of the paper is devoted to laying out these issues , and resolving them .
once that is done , the proof of finite nuclear dimension is very similar to @xcite .
here is an outline of what follows . in section [ preliminaries ]
, we establish notation and give relevant background information on nuclearity , nuclear dimension , hilbert @xmath0-modules , @xmath0-correspondences , and cuntz pimsner algebras . section [ rokhlin dimension for correspondences ]
is about the definition of rokhlin dimension for @xmath0-correspondences . the heavy lifting is contained in section [ heavy lifting ] , culminating with the proof of our main result ( theorem [ main_thm ] ) . in section [ amalgamated ]
, we observe a cute application : in certain circumstances , reduced amalgamated free products have finite nuclear dimension .
finally , in section [ classifiability ] , we show how to use work of schweizer ( @xcite ) to deduce corollary [ application ] from theorem [ main - theorem ] .
throughout all that follows , we will use the following conventions , assumptions , and notation . when we refer to an _ ideal _ of a @xmath0-algebra , we mean a closed , two - sided ideal .
if @xmath1 is a @xmath0-algebra , let @xmath12 , @xmath13 , @xmath14 , and @xmath15 denote the positive cone , unit ball , multiplier algebra , and center , respectively , of @xmath1 .
if @xmath16 are self - adjoint , let @xmath17 , @xmath18 , and @xmath19 mean @xmath20 , @xmath21 , and @xmath22 , respectively . if @xmath23 , let @xmath24 denote the commutant of @xmath25 in @xmath1 and let @xmath26 mean that for every contraction @xmath27 , there is a @xmath28 satisfying @xmath29 .
if @xmath30 is a linear map to a @xmath0-algebra @xmath31 , we will say that @xmath32 is @xmath33-contractive if @xmath34 .
we will use the abbreviation c.p.(c . ) to mean completely positive ( and contractive ) .
we write @xmath35 for the compact operators on @xmath36 . for a @xmath0-algebra @xmath1
, we define @xmath37 and view @xmath1 as a subalgebra of @xmath38 in the canonical way ( consisting of elements represented by constant sequences ) . throughout this section
let @xmath1 and @xmath39 be @xmath0-algebras .
[ order zero maps ] a c.p .
map @xmath40 is _ order zero _ if it preserves orthogonality : for every pair of positive elements @xmath41 , @xmath42 @xmath1 has _ nuclear dimension _ at most @xmath43 , written @xmath44 , if for every finite subset @xmath45 and @xmath46 , there is a finite - dimensional @xmath0-algebra @xmath47 , a c.p.c .
map @xmath48 , and a c.p . map @xmath49 satisfying 1 .
@xmath50 for every @xmath51 , and 2 . for each @xmath52 , the restriction of @xmath53 to @xmath54 is contractive and order zero .
the following is well - known and underpins many nuclear dimension computations in the literature .
[ nucdimtool ] fix @xmath55 .
let @xmath1 be a @xmath0-algebra .
then @xmath56 if , for every finite set @xmath57 and @xmath58 , there exists a @xmath0-algebra @xmath31 of nuclear dimension at most @xmath43 and c.p .
maps @xmath59 such that @xmath53 is c.p.c .
, @xmath60 is a sum of @xmath61 c.p.c .
order zero maps , and @xmath62 for @xmath63 . by ( * ? ? ?
* proposition 2.5 ) , we only need to show that the inclusion of @xmath1 into @xmath38 has nuclear dimension at most @xmath64 to this end , let @xmath57 be finite and let @xmath58 .
find @xmath65 , and @xmath60 as in the hypotheses , for the finite set @xmath66 and with @xmath67 in place of @xmath33 .
since @xmath31 has nuclear dimension at most @xmath61 , we can factor the identity map on @xmath31 , up to @xmath68 on @xmath69 , as @xmath70 such that @xmath71 is c.p.c . and
@xmath9 is a sum of @xmath72 c.p.c .
order zero maps . then the inclusion @xmath73 factors , up to @xmath33 on @xmath66 , as @xmath74 where @xmath75 is c.p.c . , and @xmath76 decomposes as a sum of @xmath77 c.p.c .
order zero maps . if @xmath78 is an ideal , then by ( * ? ? ?
* proposition 2.9 ) we have @xmath79 however a slight modification of the proof of ( * ? ? ?
* proposition 5.1 ) implies the following statement .
[ propextension ] if @xmath80 has a quasicentral approximate unit consisting of projections , then @xmath81 . throughout this section
let @xmath1 be a unital @xmath0-algebra .
we give a brief overview of @xmath0-modules ( all of which are assumed to be over @xmath1 ) , @xmath0-correspondences , and cuntz pimsner algebras . for comprehensive treatments , see @xcite .
let @xmath2 be a hilbert @xmath0-module .
we say @xmath2 is _ full _ if the set @xmath82 is dense in @xmath1.we say @xmath2 is _ free _ if it has an orthonormal set of generators ; it is _ finitely generated projective _ if it is an orthogonal direct summand in a finitely generated free module .
if @xmath83 is another @xmath0-module , we denote by @xmath84 and @xmath85 the adjointable and compact operators from @xmath2 to @xmath83 , respectively .
@xmath85 is the closed span of operators @xmath86 ( over @xmath87 ) , where @xmath88 in the case @xmath89 , we write @xmath90 and @xmath91 . besides the operator norm topology , there is another natural topology on @xmath84 : a sequence @xmath92 converges _ strictly _ to @xmath93 if @xmath94 and @xmath95 for every @xmath96 and @xmath97 .
if @xmath98 is a collection of @xmath0-modules , their direct sum @xmath99 , defined as @xmath100 is also a @xmath0-module .
if @xmath2 is a free @xmath0-module whose orthonormal generators are indexed by a set @xmath101 , then @xmath102 .
we will need the following important theorem of kasparov .
[ kasparov s stabilization theorem ] if @xmath2 is a countably generated , then there is a countably generated free @xmath0-module @xmath103 satisfying @xmath104 kasparov s stabilization theorem implies @xmath91 is isomorphic to a hereditary subalgebra of @xmath105 . by (
* proposition 2.5 ) , this implies @xmath106 we say @xmath2 is a _ @xmath0-correspondence _ ( or simply a correspondence ) if there is a unital injective * -homomorphism @xmath107 .
the map @xmath108 is called the left action of @xmath1 on @xmath2 and we will write @xmath109 as @xmath110 .
the simplest example is the _ identity correspondence _ , which is the @xmath0-module @xmath1 ( over @xmath1 ) with the left action given by left multiplication .
if @xmath98 is a collection of correspondences , their direct sum @xmath99 is also a correspondence with the left action given by @xmath111 .
we say two correspondences @xmath2 and @xmath83 are _ unitarily equivalent _
( and write @xmath112 ) if there exists an adjointable map @xmath113 such that 1 . @xmath114 for every @xmath96 and @xmath115 , and 2 .
@xmath116 , and @xmath117 . the algebraic tensor product @xmath118 of two correspondences @xmath2 and
@xmath83 is naturally a right @xmath1-module with an @xmath1-valued semi - inner product given by @xmath119 denote by @xmath120 the @xmath0-module obtained from @xmath121 by separation and completion . in @xmath122 ,
the following identity holds @xmath123 there is an injective * -homomorphism @xmath124 given by @xmath125 , where @xmath126 ( and a * -homomorphism @xmath127 given by @xmath128 ) .
in particular @xmath120 is a correspondence , called the _
interior tensor product _ of @xmath2 and @xmath83 .
elements of the form @xmath129 are called _ elementary tensors_. if @xmath130 is a correspondence over a @xmath0-algebra @xmath1 , define @xmath131 and one can easily check that this is a correspondence over the sequence algebra @xmath38 in the obvious way .
it contains @xmath130 as the subset consisting of elements with constant sequence representatives .
in particular , the product ( in either order ) of an element of @xmath38 and an element of @xmath130 makes sense as an element of this @xmath132 . for a correspondence @xmath130 over a @xmath0-algebra @xmath1 , a _ representation _ of @xmath2 on a @xmath0-algebra @xmath31
is a pair @xmath133 consisting of a * -homomorphism @xmath134 and a linear map @xmath135 satisfying @xmath136 and @xmath137 for every @xmath138 and @xmath139 .
denote by @xmath140 the @xmath0-subalgebra of @xmath31 generated by @xmath141 and @xmath142 .
we say a representation @xmath133 _ admits a gauge action _ if there is an action @xmath71 of @xmath143 on @xmath140 such that @xmath144 and @xmath145 for all @xmath115 and @xmath146 .
we now briefly review the construction of cuntz pimsner algebras .
these were first defined by pimsner in @xcite .
note that @xcite contains the proofs of theorems [ toeplitz - properties ] and [ cuntz - universal ] , and proposition [ useful_cuntz ] .
see also ( * ? ? ?
* section 4.6 ) for a single correspondence @xmath2 , set @xmath147 and @xmath148 .
the _ full fock space _ and _ @xmath149-cutoff fock space _ over @xmath2 are the correspondences defined by @xmath150 we say an elementary tensor @xmath151 has _ length @xmath152 _ and write @xmath153 .
when @xmath130 is a free module with orthonormal basis @xmath154 , we set @xmath155 equal to the set of all elementary tensors in the @xmath156 , and @xmath157 the left action of @xmath1 on @xmath158 is given by @xmath159 .
moreover for each @xmath160 we define the _ creation _ operator @xmath161 by @xmath162 and @xmath163 . if @xmath164 set @xmath165 , and if @xmath166 set @xmath167 .
note that for an elementary tensor @xmath168 , @xmath169 using rank - one operators @xmath170 associated to elementary tensors in @xmath158 , we have the identity @xmath171 where convergence is understood to be strict .
[ toeplitz - pimsner algebra ] let @xmath2 be a correspondence over @xmath1 .
the _ toeplitz
pimsner algebra _
@xmath172 is the @xmath0-subalgebra of @xmath173 generated by @xmath1 and @xmath174 .
[ toeplitz - properties ] let @xmath172 be the toeplitz
pimsner algebra of a correspondence @xmath2 over @xmath1 . 1 . for every @xmath175 , @xmath138 , and @xmath139 , the creation operators satisfy @xmath176 in particular ,
@xmath177 2 .
( @xcite ) the toeplitz
pimsner algebra @xmath178 is the unique @xmath0-algebra ( up to isomorphism ) generated by a representation @xmath133 of @xmath130 , such that @xmath179 and which admits a gauge action .
the toeplitz
pimsner algebra is too large for many purposes , so we define the cuntz
pimsner algebra @xmath11 to be a natural quotient of @xmath172 .
denote by @xmath180 the @xmath0-subalgebra of @xmath173 generated by @xmath181 the multiplier algebra @xmath182 can be identified with all @xmath183 satisfying both @xmath184 and @xmath185 . in particular , there is there is an inclusion @xmath186 .
( however , note that @xmath172 may not contain @xmath187 . ) [ cpalgebra ] the _ cuntz pimsner algebra _
@xmath11 is the @xmath0-algebra @xmath188 , where @xmath189 is the quotient map .
we denote by @xmath190 the image of the creation operator @xmath191 under @xmath192 .
here is another description of @xmath11 .
let @xmath193 .
since @xmath194 is an ideal in @xmath1 and @xmath195 is a @xmath173-invariant subcorrespondence of @xmath158 , we can conclude @xmath196 is an ideal in @xmath173 .
[ useful_cuntz ] @xmath197 , and in particular @xmath198 . in other words , @xmath199 [ cuntz - universal ] let @xmath11 be the cuntz
pimsner algebra of a correspondence @xmath2 over @xmath1 . for every @xmath200 , and @xmath139 ,
the following identity holds : @xmath201 [ cuntz = pimsner ] if @xmath2 is a finitely generated projective correspondence , then @xmath202 ( see @xcite ) and hence @xmath203 .
since @xmath1 is unital , @xmath204 .
this shows @xmath172 contains all of @xmath205 and @xmath206 . at the other extreme ,
if @xmath207 , the kernel of @xmath192 is trivial and hence there is a * -isomorphism @xmath208 sending @xmath209 to @xmath209 and @xmath190 to @xmath191 .
here is our definition of rokhlin dimension for @xmath0-correspondences .
[ rokhlin - dimension - def ] let @xmath1 be a separable @xmath0-algebra and let @xmath2 be a countably generated correspondence over @xmath1 .
we say that @xmath2 has _
rokhlin dimension at most @xmath210 _ , @xmath211 , if any @xmath212 , there exist positive contractions @xmath213 satisfying 1 .
@xmath214 for all @xmath215 and all @xmath216 , 2 .
@xmath217 , and 3 .
@xmath218 in @xmath132 , for all @xmath152 , @xmath215 , and @xmath219 .
[ rokhlin - def - rem ] ( i ) one can of course reformulate this definition without using the sequence algebra @xmath38 and the sequence correspondence @xmath132 .
namely , we have @xmath220 if and only if , for any @xmath46 , any @xmath212 , any finite set @xmath45 , and any finite set @xmath221 , there exist positive contractions @xmath222 satisfying 1 .
@xmath223 when @xmath216 and all @xmath215 .
2 . @xmath224 .
@xmath225 for all @xmath152 , @xmath215 , and @xmath226 .
4 . @xmath227\rvert < \epsilon$ ] for all @xmath228 and @xmath51 .
\(ii ) if @xmath1 is a @xmath0-algebra and @xmath229 , one may define a correspondence @xmath230 over @xmath1 as the singly generated @xmath0-module @xmath1 with with the left action given by @xmath231 .
there is a canonical isomorphism @xmath232 ( this isomorphism fixes @xmath1 and sends @xmath233 to the canonical unitary of the crossed product ) .
our definition of rokhlin dimension of @xmath0-correspondences is designed to ( almost ) coincide with the rokhlin dimension of @xmath9 ( with single towers ) as defined by hirshberg , winter , and zacharias in ( * ? ? ?
* definition 2.3(c ) ) .
specifically , we ask for single towers all of height @xmath234 ( whereas their definition of @xmath235 asks that each colour has a single tower of height either @xmath236 or @xmath234 ) . in ( * ?
* proposition 2.8 and remark 2.9 ) , they show that , up to a possible factor of @xmath237 , their rokhlin dimension coincides with the version with single towers all of height @xmath234 . the same argument applies to variants on the definition of rokhlin dimension for correspondences , and as such , we have chosen to work with the simplest version of rokhlin dimension .
\(iii ) we can simultaneously express condition ( 3 ) and the requirement that the rokhlin contractions commute with @xmath1 by asking that , for any elementary tensor @xmath238 , @xmath239
in ( * ? ? ? * theorem 4.1 ) , it is shown that for a @xmath0-algebra @xmath1 with finite nuclear dimension and an automorphism @xmath240 of finite rokhlin dimension , the crossed product @xmath241 has finite nuclear dimension . in this section
we generalize this result to correspondences of finite rokhlin dimension ( this is truly a generalization , see remark [ rokhlin - def - rem ] ( i ) ) , subject to a technical condition which is satisfied , for example , by correspondences which are finitely generated and projective as hilbert @xmath0-modules .
one can recast the argument used to prove ( * ? ? ?
* theorem 4.1 ) in terms of the fock - space representation of @xmath242 , where @xmath230 is as in remark [ rokhlin - def - rem ] ( i ) .
specifically , the argument makes use of outgoing maps from @xmath243 to a compression of the fock space representation ; the range of these maps land in a subalgebra of @xmath244 which , in this case , is isomorphic to some @xmath245 .
it is possible to then define incoming maps @xmath246 , using compressions by row vectors corresponding to rokhlin towers .
our argument for general correspondences is based on this outline ; however , there are two significant technical differences .
first , the codomain of the outgoing map will generally not be a matrix algebra over @xmath1 , and so further input is needed to get a ( uniform ) bound on its nuclear dimension .
second , there need to be suitable replacements for the row vector compressions used to construct the incoming maps . in this section
, we deal with these technicalities .
let @xmath2 be a countably generated correspondence over @xmath1 . for each @xmath247 , form @xmath248 set @xmath249
this is a @xmath0-subalgebra of @xmath250 .
evidently , @xmath251 is an ideal of @xmath252 .
[ quasi_hypothesis ] suppose that for every @xmath212 , there is an approximate unit consisting of projections in @xmath253 that are quasicentral in @xmath252 .
then for every @xmath212 we have @xmath254 by induction in @xmath236 .
it is not hard to see that @xmath255 .
suppose the result holds for @xmath256 for @xmath257 .
set @xmath258 it s clear that @xmath259 and that @xmath260 this shows that @xmath261 is an extension of ( an algebra isomorphic to ) @xmath252 by the compacts @xmath262 .
the result follows by induction , proposition [ propextension ] , and by the inductive hypothesis and , both of these algebras have nuclear dimension at most @xmath263 . by the hypotheses of this lemma
, we may apply proposition [ propextension ] to conclude that the nuclear dimension of @xmath264 is also at most @xmath263 .
[ qd - examples ] ( i ) if @xmath2 is a finitely generated projective correspondence , then so is @xmath265 for any @xmath266 ( see ( * ? ? ?
* proposition 4.7 ) ) .
hence , @xmath253 is unital for any @xmath212 and we get @xmath267 thus , such @xmath130 does satisfy the condition of lemma [ quasi_hypothesis ] ( i.e. , for every @xmath236 , there is an approximate unit of projections in @xmath268 which is quasicentral in @xmath269 ) .
\(ii ) let @xmath83 be countably generated free hilbert @xmath1-module with orthonormal basis @xmath270 .
let @xmath271 be a sequence of automorphisms of @xmath1 , and define a correspondence by the left action @xmath272 for @xmath273 .
this correspondence also satisfies the hypothesis of lemma [ quasi_hypothesis ] . for @xmath274 , set @xmath275 and note that @xmath276 .
let @xmath277 denote the finite subset of @xmath278 consisting of elementary tensors of length @xmath279 in @xmath280 .
define the finite rank projection @xmath281 .
this projection is quasicentral in @xmath282 , since for @xmath283 and @xmath284 , for sufficiently large @xmath43 ( namely , when @xmath43 is greater than all indices of basis elements appearing in @xmath285 and @xmath286 ) , we have @xmath287 this section works extensively with countably generated free correspondences : first we use the structure of such a correspondence @xmath288 to define incoming maps @xmath289 , and then we use kasparov s stabilization theorem to apply such incoming maps to general countably generated correspondences . throughout , @xmath288
will generally be a countably generated free correspondence , while @xmath130 will be an arbitrary ( countably generated ) hilbert module or correspondence . for a countably generated free correspondence @xmath288 over @xmath1 with orthonormal basis @xmath270 , recall that @xmath290 denotes the set of elementary tensors in this generating set , while ( from ) @xmath291 and @xmath278 denote the subsets of elementary tensors of length @xmath152 and @xmath279 respectively . as hilbert @xmath1-modules , we have @xmath292 [ i - love - lamp ] let @xmath1 be a @xmath0-algebra , let @xmath293 be a hilbert @xmath1-module , and let @xmath101 be an index set .
there is an inclusion @xmath294 .
more specifically , an operator @xmath295 acts adjointably on @xmath296 via @xmath297 where @xmath298 is the canonical orthonormal basis for @xmath299 . the map
@xmath300 implements an isomorphism between @xmath301 and @xmath293 .
let us show that the map @xmath302 extends to an isomorphism between @xmath303 and @xmath296 .
we have @xmath304 so that @xmath61 can be extended , as a @xmath305-preserving linear functional , to the tensor product @xmath303 .
moreover , the image of @xmath61 contains the dense set consisting of finitely supported elements in @xmath296 , so that @xmath61 is a unitary operator .
the result follows since @xmath306 is a * -isomorphism between @xmath307 and @xmath308 , and composing this with the natural embedding @xmath309 yields a map satisfying .
[ techcortwo ] let @xmath1 be a @xmath0-algebra , let @xmath288 be a countably generated free correspondence with orthonormal basis @xmath310 , and define @xmath278 by . for each @xmath247
, there is an inclusion @xmath311 that sends @xmath312 ( where @xmath283 , @xmath313 , and @xmath314 ) @xmath315 and @xmath209 ( as an operator in @xmath316 ) to the operator @xmath317 applying lemma [ i - love - lamp ] with @xmath318 and @xmath319 yields an embedding @xmath320 such that for @xmath321 and @xmath322 , @xmath323 we use the canonical identification of @xmath324 with @xmath325 to view @xmath326 as an inclusion of @xmath327 to @xmath328 .
this identification takes an elementary tensor @xmath329 to the orthonormal basis element @xmath330 . since @xmath313 acts on @xmath324 by sending @xmath329 to @xmath331 , follows immediately . for , using and , first note that for @xmath329 , @xmath332 for @xmath333 where @xmath334 and for @xmath335 , @xmath336 putting this together , we have @xmath337 as required .
[ rows - vectors ] let @xmath1 be a @xmath0-algebra , let @xmath293 be a hilbert @xmath1-module , and let @xmath101 be a countable index set . let @xmath338 be a collection of isometries in @xmath339 with orthogonal ranges such that @xmath340 converges strictly in @xmath339 .
then the map @xmath341_i:\bigoplus_i\mathscr{h } \rightarrow \mathscr{h}$ ] given by @xmath342 is an adjointable operator with adjoint given by @xmath343_i^*(x ) = ( t_i^*x)_{i \in i}.\ ] ] moreover , @xmath341_i$ ] is an isometry .
the hypothesis that @xmath340 converges strictly implies that the formula for @xmath341_i^*(x)$ ] does define an element of @xmath344 , and therefore this formula produces a well - defined map @xmath345 . for an indexed family @xmath346 of pairwise orthogonal elements of @xmath130 , note that @xmath347 converges to an element @xmath348 if and only if @xmath346 represents an element @xmath168 of @xmath344 , and in this case , @xmath349 . since the @xmath350 have orthogonal ranges , it follows from these facts that @xmath341_i$ ] is a well - defined isometry .
it is an easy calculation to see that the formula for @xmath341_i^*$ ] does indeed provide an adjoint to @xmath341_i$ ] .
the assumption that @xmath340 converges strictly is not automatic ; here is an example .
let @xmath351 and let @xmath352 ; write an element of @xmath130 as @xmath353 where for each @xmath354 , @xmath355 converges ( and is uniformly bounded in @xmath354 ) . for each @xmath356 , pick an injective map @xmath357 such that @xmath358 .
for each @xmath284 , define @xmath359 by @xmath360 where @xmath361 using the fact that the @xmath362 are injective , one easily computes @xmath363 so that each @xmath364 is isometric
. injectivity of the @xmath362 also implies that the @xmath364 have pairwise orthogonal ranges . to see that @xmath365 does not converge strictly ,
let us check that @xmath366 does not converge ( in @xmath130 ) .
note that @xmath367 , and that for @xmath368 , @xmath369 where @xmath370 by our choice of @xmath362 , we have @xmath371 if and only if @xmath372 . therefore , @xmath373 ( viewing @xmath374 as an element of @xmath375 )
. the sequence @xmath376 does not converge ( in norm ) in @xmath1 , so that @xmath366 does not converge in @xmath130 .
[ techcorone ] let @xmath1 be a @xmath0-algebra , let @xmath288 be a countably generated free correspondence with orthonormal basis @xmath310 , and define @xmath377 by .
for any @xmath266 , there is an isometry @xmath378_{w_k } \in \mathbb{b}(\bigoplus_{w_k}\mathcal{f}(\mathcal{k } ) , \mathcal{f}(\mathcal{k}))$ ] given by @xmath379 note that @xmath380 is a family of isometries with pairwise orthogonal images , and that @xmath381 converges strictly to @xmath382 in @xmath383 . thus setting @xmath384 and @xmath385 , the hypotheses of lemma [ rows - vectors ] are satisfied
; this lemma shows that @xmath386_{w_k}$ ] is a well - defined isometry . for each @xmath247 and @xmath387 , we can regard @xmath378_{w_k}$ ] as being an element in @xmath388 by identifying @xmath389 with @xmath390 , and defining @xmath378_{w_k}$ ] to be zero on @xmath391 for @xmath392 .
recall from definition [ cpalgebra ] that , for @xmath393 , @xmath190 denotes the image of @xmath191 in the cuntz
pimsner algebra @xmath394 .
[ maps ] let @xmath1 be a @xmath0-algebra , let @xmath288 be a countably generated free correspondence with orthonormal basis @xmath310 , and define @xmath278 by .
let @xmath247 and let @xmath395 be an tuple of positive contractions in @xmath1 .
set @xmath396_{k } \in \mathbb{b}(\bigoplus_{w_{<p}}\mathcal{f}(\mathcal{k } ) , \mathcal{f}(\mathcal{k})).\ ] ] define @xmath397 to be the following composition : @xmath398 where the first map is the inclusion given by corollary [ techcortwo ] .
let @xmath399 denote the quotient map .
then @xmath400 , and so there exists a c.p .
map @xmath401 this map satisfies @xmath402 for elementary tensors @xmath403 in @xmath404 and @xmath284 such that @xmath405 . for @xmath115 , @xmath406 , and @xmath266 such that @xmath407 , and @xmath408 , we have @xmath409 note that for a basis element @xmath410 , @xmath411 and by linearity , this formula continues to hold for all @xmath412 .
putting these together , we find @xmath413 define @xmath414 , and we see that holds for @xmath415 where @xmath406 . by density , holds for all elementary tensors @xmath403 , and the image of @xmath416 is contained in @xmath394 ( i.e. , the image of @xmath417 is contained in @xmath418 ) .
[ moremaps ] let @xmath1 be a @xmath0-algebra , let @xmath288 be a countably generated free correspondence with orthonormal basis @xmath310 , and define @xmath278 by .
let @xmath419 , let @xmath420 be the maximum value of @xmath421 , @xmath422 , and define @xmath423 by . 1 .
regarding @xmath424 as an element of @xmath425 , we have @xmath426 2 .
let @xmath427 and @xmath428 be elementary tensors in @xmath404 satisfying @xmath429 and @xmath430 . if @xmath284 is such that @xmath431 , then @xmath432\rvert < 2(p^2 + 2)\delta.\ ] ] ( 1 ) : for @xmath433 , @xmath434_{w_k}^*g_k^2[t_\eta]_{w_k}(x_\zeta)_{\zeta\in w_{<p } } & = [ t_\eta]_{w_k}^*\sum_{\zeta \in w_k } g_k\cdot \zeta \otimes x_\zeta \\ & = ( \chi_{w_k}(\zeta ) g_k^2\cdot x_\zeta)_{\zeta \in w_{<p}},\end{aligned}\ ] ] and therefore , @xmath435_{w_k}^*g_k^2[t_\eta]_{w_k } \\ &
\approx_{p^2\delta } \sum_{k , k ' } [ t_\eta]_{w_k}^*g_kg_{k'}[t_\eta]_{w_{k ' } } \\ & = r_{\mathcal g}^*r_{\mathcal g}.\end{aligned}\ ] ] where @xmath436 and @xmath437 .
( 2 ) : using ( 1 ) we have @xmath438 let @xmath2 be a countably generated correspondence over @xmath1 . by kasparov s stabilization theorem ( theorem [ kasparov s stabilization theorem ] ) , there is a countably generated free hilbert @xmath1-module @xmath103 such that their direct sum @xmath439 is free . choose a left action of @xmath1 on @xmath103 such that @xmath440 ( one may take , for example , the canonical left action of @xmath1 on @xmath441 ) .
the diagonal action of @xmath1 turns @xmath83 into a correspondence .
the orthogonal projection @xmath442 onto @xmath443 commutes with the image of the left action of @xmath1 on @xmath83 , so for @xmath444 the map @xmath445 is an orthogonal projection in @xmath446 .
let @xmath1 be a @xmath0-algebra , let @xmath130 be a countably generated correspondence over @xmath1 , and let @xmath288 and @xmath447 be as above . allowing @xmath448 to mean the identity map on @xmath1 , we have 1 .
@xmath449 , 2 .
@xmath450 , and 3 .
@xmath451 . for the first part , the map @xmath452 implements an isomorphism between the @xmath152-fold tensor product @xmath265 and the image of @xmath453 in @xmath454 .
the second part follows easily from the first since the sums are finite .
lastly , it is clear that the series @xmath455 converges strictly to a projection in @xmath383 with the desired property .
denote the projections @xmath456 and @xmath455 given above by @xmath457 and @xmath458 , respectively .
if @xmath459 is an elementary tensor , we say that @xmath428 is an _
@xmath2-elementary tensor _
if @xmath460 .
equivalently , @xmath428 is @xmath2-elementary if @xmath461 .
[ sumuniversecuntz ] let @xmath133 be the representation of @xmath2 in @xmath462 given by @xmath463 and @xmath464 .
then there is a * -isomorphism @xmath465 sending @xmath209 to @xmath209 and @xmath466 to @xmath191 .
consider first the representation @xmath467 of @xmath2 in @xmath468 given by @xmath469 and @xmath470 .
it is clear that @xmath471 if we restrict the gauge action on @xmath468 to @xmath140 , theorem [ toeplitz - properties ] ( 2 ) gives a * -isomorphism @xmath472 sending @xmath209 to @xmath209 and @xmath473 to @xmath191 .
now , since the left action of @xmath1 on @xmath103 was defined so that @xmath474 , each @xmath475 differs from any finite sum of rank - one operators by at least @xmath476 .
therefore , @xmath477 and so by remark [ cuntz = pimsner ] there is a * -isomorphism @xmath478 sending @xmath209 to @xmath209 and @xmath479 to @xmath480 .
moreover , it s clear that @xmath481 so taking @xmath482 completes the proof .
here is a diagram illustrating the maps involved .
\(m ) [ matrix of math nodes , row sep=4em , column sep=9em ] ( a , ) & c^ * ( , ) ( ) + ( ) & c^ * ( , ) ( ) + ; ( m-1 - 1 ) edge node[auto ] @xmath483 ( m-1 - 2 ) ( m-1 - 1 ) edge node[auto , swap ] @xmath484 ( m-2 - 1 ) ( m-1 - 2 ) edge node[auto ] @xmath485 ( m-2 - 2 ) ( m-2 - 2 ) edge node[auto ] @xmath486 ( m-2 - 1 ) ( m-1 - 1 ) edge node[descr ] @xmath487 ( m-2 - 2 ) ; [ embed ] let @xmath1 be a @xmath0-algebra , let @xmath130 be a countably generated correspondence over @xmath1 , and let @xmath288 be as above . let @xmath488 be the @xmath0-subalgebra of @xmath282 generated by elements of the form @xmath489 where @xmath490 and @xmath168 are @xmath2-elementary tensors
. then the map @xmath491 extends to an isomorphism @xmath492 . in particular
if for each @xmath236 , there is an approximate unit of projections in @xmath268 which is quasicentral in @xmath269 , then @xmath493 .
the map @xmath494 sending @xmath209 to @xmath209 and @xmath495 to @xmath496 extends to a unitary from @xmath497 to @xmath498 .
let @xmath499 satisfy @xmath500 and let @xmath501 be elementary tensors of lengths @xmath502 and @xmath152 , respectively .
we have @xmath503 so @xmath504 is implemented by unitary conjugation and the result follows .
[ spock_cor ] let @xmath1 be a @xmath0-algebra , let @xmath130 be a countably generated correspondence over @xmath1 , and let @xmath288 be as above .
let @xmath419 be a tuple of positive contractions in @xmath1 .
there is a c.p .
map @xmath505 such that , for elementary tensors @xmath403 in @xmath497 and @xmath152 such that @xmath506 , @xmath507 and the following commutes \(m ) [ matrix of math nodes , row sep=2em , column sep=8em , text height=1.5ex , text depth=0.25ex ] d_p ( ) & d_p^0()d_p ( ) + ( ) & c^ * ( , ) ( ) + ; ( m-1 - 1 ) edgenode[auto ] @xmath504 ( lemma [ embed ] ) ( m-1 - 2 ) ( m-1 - 1 ) edgenode[auto ] @xmath508 ( m-2 - 1 ) ( m-1 - 2 ) edgenode[auto ] @xmath509 ( lemma [ maps ] ) ( m-2 - 2 ) ( m-2 - 2 ) edgenode[auto ] @xmath510 ( prop .
[ sumuniversecuntz ] ) ( m-2 - 1 ) ; define @xmath511 by lemma [ maps ] and @xmath512 from lemma [ embed ] .
let @xmath513 and @xmath514 .
we compute @xmath515 and we see that the result is in @xmath140 where @xmath133 is as in proposition [ sumuniversecuntz ] .
hence , the image of @xmath516 is contained in @xmath140 and using @xmath517 defined by proposition [ sumuniversecuntz ] , we may define @xmath518 we further compute @xmath519 as required .
let @xmath1 be a @xmath0-algebra , let @xmath130 be a countably generated correspondence over @xmath1 . consider a tuple @xmath520 of positive contractions from @xmath38 .
we may lift this to a sequence @xmath521 of tuples of positive contractions from @xmath1 . using this lift and corollary [ spock_cor ] ,
we define @xmath522 note that this is independent of the choice of the lift @xmath521 .
[ lulu ] let @xmath1 be a separable @xmath0-algebra , let @xmath130 be a countably generated correspondence over @xmath1 , and let @xmath523 . if @xmath419 is a tuple of orthogonal positive contractions in @xmath38 satisfying @xmath524 for all @xmath152 and all elementary tensors @xmath525 then the c.p .
map @xmath508 from is contractive and order zero .
analogously to , define @xmath526 and @xmath527 .
we have @xmath528 by lemma [ moremaps ] ( 1 ) and the hypothesis , @xmath529\rvert \to 0\ ] ] for all @xmath130-elementary tensors @xmath530 . by the definition of @xmath504 ( from lemma [ embed ] )
, this implies that @xmath531 \rvert \to 0\ ] ] for all @xmath532 . by the formula for @xmath533 ( from lemma [ maps ] )
, it follows that @xmath417 is order zero .
consequently , @xmath534 , and therefore also @xmath535 , is order zero . to see that @xmath535 is contractive
, we have by lemma [ moremaps ] ( 2 ) that @xmath536 where @xmath537 .
since the @xmath538 are orthogonal and contractive , @xmath539 is contractive .
it follows from the formula for @xmath533 that @xmath417 is contractive .
hence , so is @xmath535 .
we have one technical lemma before the main result . for @xmath523 and @xmath540 , set @xmath541\ ] ] + .
[ bumplemma ] let @xmath236 be an odd integer .
if @xmath542 are positive contractions in a @xmath0-algebra @xmath1 , then for any @xmath543 , @xmath544 it s easy to verify that @xmath545 so that @xmath546 moreover , @xmath547 so the result follows .
we now have the ingredients to prove the main result . [ main_thm ]
suppose that @xmath2 is a countably generated @xmath0-correspondence over a separable unital @xmath0-algebra @xmath1 .
assume further that for every @xmath212 , there is an approximate unit consisting of projections in @xmath253 that are quasicentral in @xmath252 .
then @xmath10 example [ qd - examples ] gives two important examples of classes of correspondences which satisfy the technical hypothesis of this theorem , namely , that for every @xmath236 , there is an approximate unit of projections in @xmath268 which is quasicentral in @xmath269 .
if either @xmath263 or @xmath548 there is nothing to show . otherwise , let @xmath549 and @xmath550 .
we will use lemma [ nucdimtool ] with @xmath551 .
therefore let @xmath66 be a finite subset of @xmath172 and fix @xmath46 .
noting , we may assume without loss of generality that @xmath66 consists of elements of the form @xmath552 where @xmath403 are elementary tensors .
pick @xmath553 such that every element of @xmath66 is @xmath552 for some elementary tensors @xmath554 .
let @xmath247 be large enough so that @xmath555 and let @xmath556 be the projection onto the @xmath149-cutoff fock space @xmath497 . for @xmath557 ,
use @xmath558 as defined in ; for convenience , we set @xmath559 for @xmath560 . using these , set @xmath561 and let @xmath562 be compression by @xmath563 ; that is , for elementary tensors @xmath564 , @xmath565 ( here it is convenient that @xmath559 for @xmath560 ) .
it is not hard to see that @xmath566 is contained in @xmath269 .
find rokhlin contractions @xmath567 in @xmath38 satisfying 1 .
@xmath214 for all @xmath215 and @xmath568 , 2 .
@xmath569 , 3 .
@xmath570 for all @xmath228 , and all elementary tensors @xmath238 .
note that ( 3 ) implies that @xmath571 for all @xmath572 , which in turn implies that @xmath573 for any continuous function @xmath574)$ ] .
in particular , @xmath575 for @xmath576 , set @xmath577 two tuples of orthogonal positive elements in @xmath38 .
define @xmath578 and @xmath579 as in corollary [ spock_cor ] ; using , corollary [ lulu ] tells us that these maps are c.p.c . and order zero .
to simplify notation , write @xmath580 and @xmath581 for @xmath578 and @xmath579 , respectively .
thus we have the following diagram .
\(m ) [ matrix of math nodes , row sep=2em , column sep=8em , text height=1.5ex , text depth=0.25ex ] t(h ) & & t(h ) _ + & d_p(h ) & + ; ( m-1 - 1 ) edgenode[auto ] ( m-1 - 3 ) ( m-1 - 1 ) edgenode[auto , swap ] @xmath60 ( m-2 - 2 ) ( m-2 - 2 ) edgenode[auto , swap ] @xmath582 ( m-1 - 3 ) ; by lemma [ quasi_hypothesis ] , @xmath269 has nuclear dimension at most @xmath43 .
we shall show that this diagram commutes , up to @xmath33 on @xmath66 . as @xmath60 is c.p.c .
and each @xmath583 is overtly a sum of @xmath584 c.p.c .
order zero maps , this will finish verifying the hypotheses of lemma [ nucdimtool ] , and thus finish the proof .
consider an element of @xmath66 , necessarily of the form @xmath552 , @xmath554 .
for the moment , assume that @xmath585 .
applying @xmath60 to @xmath552 , using , and noting that @xmath586 is within @xmath587 of @xmath588 for @xmath589 , we have @xmath590 where the error estimate in the last line is a maximum ( rather than a sum ) because of orthogonality . next , applying @xmath580 to @xmath591 , we get @xmath592 and likewise , @xmath593 summing these terms , we obtain @xmath594 by , this yields @xmath595 as required .
a nearly identical argument shows the same estimate in the case @xmath596 .
[ main - theorem_cor ] under the same hypotheses as the previous theorem , @xmath597 this follows from the fact that @xmath11 is a quotient of @xmath172 by the ideal @xmath598 , and by ( * ? ? ?
* proposition 2.3 ) .
we now use the results given in the previous sections to deduce finite nuclear dimension of a certain class of reduced amalgamated free products . for a more comprehensive treatment of free products ,
see the monograph @xcite .
we start with a well - known example .
[ free - group ] there is a * -isomorphism @xmath599 here the free product is being taken with respect to the state @xmath600 on @xmath601 .
the free group @xmath602 on two generators is not amenable , so the reduced group @xmath0-algebra is not nuclear and in particular has infinite nuclear dimension .
what this shows is that unlike other canonical @xmath0-constructions , finite nuclear dimension is not in general preserved under the reduced amalgamated free product construction , even in the abelian case .
however , the next result by speicher from @xcite shows that there are exceptions .
[ speicher ] let @xmath603 be correspondences over @xmath1
. then @xmath604 [ 2-main - theorem ] let @xmath2 be a correspondence over @xmath1 and let @xmath101 be an arbitrary ( countable ) set . if @xmath605 , then @xmath606 .
the left action of @xmath1 on @xmath607 is given by @xmath608 .
let @xmath46 , @xmath247 , @xmath66 a finite subset of @xmath1 and @xmath609 a finite subset of @xmath607 .
since the finitely supported elements of @xmath344 are dense , we may assume that @xmath610 consists only of finitely supported elements ; let @xmath611 be such that each is supported on at most @xmath553 elements .
set @xmath612 equal to the set of all elements in @xmath130 which appear in some component of some element of @xmath610 ; this is a finite set .
find rokhlin contractions @xmath613 for the correspondence @xmath130 , with respect to @xmath614 .
for @xmath615 , @xmath616 thus , the @xmath617 are rokhlin contractions for @xmath344 , with respect to @xmath618 . for countably generated hilbert modules @xmath619 and @xmath620 ,
is it the case that @xmath621 in light of theorem [ main - theorem ] and proposition [ 2-main - theorem ] , we obtain the following statement . [ second - main - theorem ]
let @xmath2 be a finitely generated projective correspondence over @xmath1 .
if @xmath263 and @xmath622 are both finite , then the amalgamated free product of any finite number of copies of @xmath172 ( with respect to the canonical expectation @xmath623 ) has finite nuclear dimension . in comparison with example [ free - group ]
, the following example demonstrates an interesting consequence of theorem [ second - main - theorem ] .
if @xmath32 is a minimal homeomorphism of @xmath624 , then @xmath625 is finite . by theorem [ second - main - theorem ] ,
the nuclear dimension of @xmath626 is also finite . here
the free product is being taken with respect to the usual conditional expectation @xmath627 .
in this section we show that in the presence of finite rokhlin dimension , the algebras @xmath11 are often _ classifiable _ , by which we mean unital and simple with finite nuclear dimension and satisfying the uct ( cf .
`` satisfying the uct '' is a @xmath628-theoretic property for separable @xmath0-algebras , equivalent to being @xmath628-equivalent to an abelian @xmath0-algebra , introduced by rosenberg and schochet in @xcite . the class of algebras which satisfy the uct ( sometimes called the _ bootstrap class _ )
is closed under a number of natural operations ( see ( * ? ? ?
* and 23 ) ) , although it is unknown whether every separable nuclear @xmath0-algebra is in this class . by ( * ?
? ? * theorem 4.4 ) , the toeplitz pimsner algebra @xmath172 is @xmath628-equivalent to @xmath1 .
consequently , @xmath172 satisfies the uct whenever @xmath1 does .
since @xmath11 is a quotient of @xmath172 by the ideal @xmath629 , it suffices to prove ( by the two - out - of - three principle ) that @xmath629 satisfies the uct . to this end
observe that @xmath630 is countably generated , so kasparov s stabilization result implies @xmath631 where @xmath632 denotes morita equivalence of @xmath0-algebras . since @xmath1 is simple , so is @xmath633 .
this means every hereditary @xmath0 subalgebra @xmath634 is full , and hence satisfies @xmath635 the result follows because @xmath629 is hereditary in @xmath636 , and because the uct is preserved under morita equivalence .
assume @xmath642 and that there is some @xmath444 such that @xmath643 .
this means there is an adjointable unitary bimodule map @xmath644 .
let @xmath645 and fix @xmath46 .
it is straightforward to check that @xmath642 implies @xmath646 , so we can find a set of positive contractions @xmath647 satisfying the following three estimates : applying @xmath651 to the third estimate gives @xmath652 for every @xmath215 .
combining this with the first estimate gives @xmath653 for @xmath654 .
but this implies @xmath655 , which contradicts the second estimate when @xmath33 is sufficiently small .
thus , @xmath2 must be nonperiodic .
combining theorem [ thm : schweizer ] and the previous lemma we see that for @xmath0-correspondences with finite rokhlin dimension , @xmath11 is simple and unital whenever @xmath1 is . for simple @xmath1 , by lemma [ classif_lem_1 ] , @xmath11 satisfies the uct whenever @xmath1 does . thus applying theorem [ main_thm ] , we obtain the following corollary .
[ application2 ] if @xmath1 is classifiable , @xmath656 and for every @xmath212 there is an approximate unit consisting of projections in @xmath253 that are quasicentral in @xmath252 ( e.g. , if @xmath2 is finitely generated projective ) , then @xmath11 is classifiable .
suppose @xmath1 is a unital kirchberg algebra and @xmath657 is an automorphism with finite rokhlin dimension ( such automorphisms are generic by ( * ? ? ? * theorem 3.4 ) ) .
if @xmath2 is a countably generated free @xmath1-module with basis @xmath658 , we can define a correspondence with the left action @xmath659 .
note that @xmath660 , so remark [ cuntz = pimsner ] and the @xmath628-equivalence of @xmath1 and @xmath172 imply @xmath1 is @xmath628-equivalent to @xmath11 .
moreover , @xmath661 , so @xmath11 is a simple , separable @xmath0-algebra with finite nuclear dimension . since the creation operator associated to @xmath662 is a proper isometry , @xmath11 must be a kirchberg algebra .
invoking kirchberg
phillips classification , we conclude that @xmath663 .
we end this paper by pointing out that our work does not recapture the interesting examples of classifiable cuntz
pimsner algebras constructed by kumjian in @xcite .
indeed , it can be shown that they do nt have finite rokhlin dimension .
pimsner , _ a class of @xmath0-algebras generalizing both cuntz krieger algebras and crossed products by @xmath666 _ , free probability theory ( waterloo , on , 1995 ) ,
fields inst , commun . ,
* 12 * , american mathematical society , providence , ri , ( 1997 ) , 189 - 212 .
a. tikuisis , s. white , and w.winter , _ quasidiagonality of nuclear @xmath0-algebras _ , to appear in ann . of math .
arxiv:1509.08318 . a. tikuisis and w. winter , _ decomposition rank of @xmath669-stable @xmath670-algebras .
pde * 7 * ( 2014 ) , no .
3 , pp . 673 - 700 . d.v .
voiculescu , k.j . dykema and a. nica ,
_ free random variables , a noncommutative probability approach to free products with applications to random matrices , operator algebras and harmonic analysis on free groups_. crm monograph series , * 1 * , american mathematical society , providence , ri , ( 1992 ) . | we extend the notion of rokhlin dimension from topological dynamical systems to @xmath0-correspondences .
we show that in the presence of finite rokhlin dimension and a mild quasidiagonal - like condition ( which , for example , is automatic for finitely generated projective correspondences ) , finite nuclear dimension passes from the scalar algebra to the associated toeplitz pimsner and ( hence ) cuntz pimsner algebras . as a consequence we provide new examples of classifiable @xmath0-algebras : if @xmath1 is simple , unital , has finite nuclear dimension and satisfies the uct , then for every finitely generated projective @xmath2 with finite rokhlin dimension , the associated cuntz
pimsner algebra @xmath3 is classifiable in the sense of elliott s program . |
minority game ( mg ) @xcite a game proposed by challet and zhang under the inspiration of the el farol bar problem @xcite is a simple model showing how selfish players cooperate with each other in the absence of direct communication .
it succinctly captures the self - organizing global cooperative behavior which is ubiquitously found in many social and economic systems . in mg , @xmath0 inductive reasoning players have to choose one out of two choices independently in each turn . based only on certain commonly available global information , each player has to decide one s choice by means of his / her current best working strategy or mental model . those who end up in the minority side ( i. e. , the choice with the least number of players ) win .
although its rules are remarkably simple , mg exhibits very rich self - organized collective behavior @xcite . moreover ,
the dynamics of mg can be explained by the so - called crowd - anticrowd theory @xcite which stated that fluctuations arisen in the mg is resulted from the interaction between crowds of like - minded agents and their anti - correlated partners .
however , in the real world , people usually do not only consider the global information when they make decisions .
in fact , they may also consult the opinion of their neighbors before making decisions .
for example , it is not uncommon for people to consider both the recommendation of their peers ( local information ) and the stock price ( global information ) in deciding which stock to buy from the market .
hence , it is instructive for us to incorporate the local information into the mg model to gain more insights on this kind of social and economic systems . in the past few years
, many researchers have studied a few variations of mg with local information @xcite .
however , players can only make use of either the global or local information in each turn in these models .
in contrast , people often make their decisions according to both the global information and their local information in many social and economic phenomenon . on the other hand , quan _ et al .
_ introduced the local information in the evolutionary minority game ( emg ) @xcite .
since we want to focus on how are players affected by their local information in a _ non - evolutionary _ game , we shall not consider the emg with local information here . with the above consideration in mind
, we would like to propose a model of mg where players use both the local and global information in an unbiased manner for making decision . in section
ii , we introduce a new model called the networked minority game ( nmg ) .
it is a modified mg model in which all players can make use of not only the global information but also the local information from their neighbors that are disseminated through a network .
results of numerical simulations are presented and discussed in section iii .
lastly , we conclude by giving a brief summary of our work in section iv .
in this section , we would like to show how to construct the networked minority game ( nmg ) model . in this repeated game
, there are @xmath0 heterogeneous inductive reasoning players whose aim is to maximize one s own profit in the game . in every turn
, each player has to choose one out of @xmath1 alternatives with label @xmath2 .
the global minority choice , denoted as @xmath3 at time @xmath4 , is simply the least popular choice amongst all players in that turn .
( note that the global minority choice is the least popular choice that are chosen by a non - zero number of players ; and it is chosen randomly amongst the choices with the least non - zero number of players in case of a tie . ) the players picked the global minority choice gain one unit of wealth while all the other lose one . unlike in mg ,
not only the global information is delivered to each player as one s external information in our model . nevertheless , the global information given to players are the same in these two models . in nmg
, we call it the global history @xmath5 which is simply the @xmath1-ary string of the global minority choice of the last @xmath6 turns . the global history can only take on @xmath7 different states .
we label these states by an index @xmath8 and denote the global history @xmath9 by the index @xmath10 . besides global information
, local information is also distributed via a network to all players in nmg such that individual player receive one s local information from a different source . in this game , we arrange all the players with label @xmath11 on a ring ( see fig . [
fig : f1 ] ) where the local information of a player is based on the choices of this player and his / her nearest neighbors on the ring .
specifically , the local information given to the @xmath12th player is the so - called local history @xmath13 which is simply the @xmath1-ary string of the local minority choice of this player of the last @xmath14 turns .
( note that we do not consider the case of @xmath15 since it is equivalent to mg . ) here , the local minority choice of the @xmath12th player at time step @xmath4 , @xmath16 , refers to the least popular choice amongst this player and his / her nearest neighbor on the ring at this time step .
( in other words , @xmath16 is the least popular choice amongst @xmath17th , @xmath12th and @xmath18th player . ) however , unlike the global minority choice , the local minority choice could be an alternative that nobody chooses . in case of a tie , @xmath16
is chosen randomly amongst the choices with the least number of players .
the local history can only take on @xmath19 different states .
we again label these states by an index @xmath20 and denote the local history of the @xmath12th player @xmath21 by the index @xmath22 . by the way , it is easy to extend nmg to the case where players are connected on a different topology .
for example , we can arrange players on a dynamically random chain in which each player is connected to two randomly chosen players and all the connections between players change at every time step . in brief , each player is given a @xmath1-ary string of length @xmath23 storing both the global and local history to decide his / her choice .
players can only interact indirectly with each other through the global history @xmath10 and their local history @xmath22 .
however , how does each player make use of such external information to decide one s choice in the nmg ?
he / she does so by employing strategies to predict the next minority choice according to both the global and local information where a strategy is a map sending individual combination of global and local history ( @xmath24,@xmath25 ) to the choice @xmath26 . a strategy @xmath27
can be represented by a vector @xmath28 where @xmath29 is the minority choice predicted by the strategy @xmath27 for the input @xmath30 . in our model , each player picks @xmath31 randomly drawn strategies from the strategy space before the game commences .
( we will discuss about the strategy space in depth later on . ) just like in mg , strategies in nmg are not evolving , i. e. , players are not allowed to revise their own @xmath31 strategies during the game . at each time step , each player uses his / her own best working strategy to guess the next global minority choice . but
how does a player decide which strategy is the best ?
players use the virtual score , which is simply the hypothetical profit for using a single strategy throughout the game , to evaluate the performance of a strategy .
the strategy with the highest virtual score is considered as the best one . since
inductive reasoning players do not know whether a strategy is good or not before the game commences , we can not restrict players to have good strategies " only . as a result , all strategies in our model must be unbiased to the input , i. e. the global and local history .
thus all the strategies employed in nmg can be picked from the full strategy space which is constituted by all the possible strategies for the input .
( it is obvious that these strategies have no bias for any input . )
there is a total of @xmath32 distinct strategies in the full strategy space for our model .
nevertheless , it is very probable that a large number of strategies exist in this full strategy space even @xmath33 .
consequently , we would like to pick strategies from the reduced strategy space @xcite to enhance computational feasibility .
the reduced strategy space is only composed of strategies which are significantly different from each other .
thus it characterizes the diversity of strategies in the full strategy space and we can use it instead of the full strategy space without altering the properties and dynamics of the game .
how can we construct the reduced strategy space for nmg ? to answer this question , let us look at the reduced strategy space of mg first . for mg ,
the reduced strategy space consists of mutually uncorrelated and anti - correlated strategies only .
challet and zhang showed that the maximal reduced strategy space of the original 2-choice mg , denoted as @xmath34 for a 2-choice game with length of global history @xmath35 , is composed of @xmath36 pairs of mutually anti - correlated strategies where any two strategies from different anti - correlated strategy pairs are uncorrelated with each other @xcite .
therefore , there is a total of @xmath37 different strategies in @xmath34 .
( note that there are other smaller reduced strategy spaces consisting of less number of strategies for mg @xcite . )
it can be shown that , in general , the maximal reduced strategy space @xmath38 for @xmath1-choice mg is composed of @xmath39 ensembles of mutually anti - correlated strategies where any two strategies from different anti - correlated strategy ensembles are uncorrelated with each other @xcite .
moreover , @xmath38 consists of @xmath40 distinct strategies . for nmg
, it seems reasonable for us to define the reduced strategy space @xmath41 to be the maximal reduced strategy space of mg .
however , we find that the cooperative behavior of the players in nmg using @xmath38 and the full strategy space are greatly different from each other .
indeed , such a discrepancy is due to the bias of the strategies of @xmath38 to some of the input ( the global and local history ) in our model .
hence , we should not use @xmath38 to substitute the full strategy space in nmg .
in fact , we must define the reduced strategy space @xmath41 in a different way . for nmg ,
a strategy @xmath42 is unbiased to the input if it satisfies the following conditions : ) simply looking at the predictions of this strategy for a _ given _ global history does not give us any information about its predictions for any other global history .
simply looking at the predictions of this strategy for a _ given _ local history does not give us any information about its predictions for any other local history .
accordingly , we define the reduced strategy space of nmg as follows : @xmath43 where the @xmath12th player uses the so - called segment strategy @xmath44 to predict the global minority choice for the global history @xmath10 if @xmath45 in case of @xmath46 and respectively for the local history @xmath22 if @xmath47 in case of @xmath48 . table [ tab : t1 ] illustrates how a strategy in @xmath41 gives the prediction of the next minority choice .
it is easy to show that the number of distinct strategies in @xmath41 is either @xmath49 or @xmath50 depending on the ratio of @xmath51 .
note that we should not define @xmath41 to be given by just one of the above expressions over all range of @xmath51 .
otherwise , there is a large redundancy of strategies in @xmath41 since many strategies are very similar for such case .
( however , it does not matter if we define @xmath41 by the second expression when @xmath52 . ) indeed , we have verified that the cooperative behavior of the players in nmg using @xmath41 and the full strategy space agree well with each other ; i. e. , @xmath41 can successfully characterize the diversity of strategy in the full strategy space .
@ccc ' '' '' global history @xmath24 & local history @xmath25 & prediction @xmath29 + ' '' '' ( 0,0 ) & ( 0 ) & 1 + ' '' '' ( 0,1 ) & ( 0 ) & 1 + ' '' '' ( 1,0 ) & ( 0 ) & 1 + ' '' '' ( 1,1 ) & ( 0 ) & 1 + ' '' '' ( 0,0 ) & ( 1 ) & 0 + ' '' '' ( 0,1 ) & ( 1 ) & 0 + ' '' '' ( 1,0 ) & ( 1 ) & 1 + ' '' '' ( 1,1 ) & ( 1 ) & 1 + to investigate how well players cooperate with each other in our game , a quantity of interest is the attendance of an alternative @xmath53 which is the number of players choosing the alternative @xmath54 at time @xmath4 . for players gaining the maximum profit ,
the expectation value of the attendance of any alternative should be equal to @xmath55 .
accordingly , the variance of the attendance @xmath56 represents the loss of players in the game .
( here , @xmath57 denotes the average over time . ) hence , we would like to study @xmath58 and @xmath59 as a function of the complexity of the system for our model . which parameter can be used as a measure of the complexity of the system ?
it is the so - called control parameter @xmath60 @xcite which is the ratio of the strategy space size to the number of strategies at play . for nmg
, the control parameter @xmath60 is equal to either @xmath61 for @xmath46 or @xmath62 for @xmath48 . on the other hand
, we also want to compare the variance of the attendance with the prediction by the crowd - anticrowd theory @xcite in order to investigate the crowding effect in our model .
since the strategies used in nmg ( which are picked from @xmath41 ) are neither anti - correlated nor uncorrelated , we can not only consider the interactions of the anti - correlated strategies in the crowd - anticrowd calculation of the variance just like in mg @xcite and the multichoice mg @xcite .
in fact , each strategy used in nmg is a set of segment strategies from either @xmath63 or @xmath64 whereas different segment strategies are used for different @xmath22 or @xmath10 ( see eq .
( [ e : rss ] ) ) .
thus we can apply the crowd - anticrowd theory to estimate the variance simply by counting the crowd - anticrowd cancellation of the anti - correlated segment strategies .
accordingly , the crowd - anticrowd prediction of the variance in nmg is given as follows : @xmath65 ^ 2 \right\ } \right\rangle & \mbox{if $ m_l \leq m_g$ , } \vspace*{1 mm } \\
\left\langle \displaystyle \frac{1}{n_c } \sum_{\mathcal{s}_l \in v_{n_c}(m_l ) } \sum_{a \in \mathcal{s}_l } \left\ { \frac{1}{n_c^2 } \left [ \sum_{b \in \mathcal{s}_l \backslash \{a\ } } ( n_{a } - n_{b } ) \right]^2 \right\ } \right\rangle & \mbox{otherwise , } \end{array } \right .
\label{e : cac_var}\ ] ] where @xmath66 and @xmath67 denotes the mutually anti - correlated strategy ensembles in @xmath64 and @xmath63 respectively , and @xmath68 is the number of players making decision according to the segment strategy @xmath69 .
we should aware that the variance of attendance for different alternatives must equal when averaged over time and initial choice of strategies since there is no bias for any alternative in our game .
in all the simulations reported in this paper , each set of data was recorded from @xmath70 independent runs . in each run
, we run 10000 steps starting from initialization before making any measurements .
we have checked that it is already enough for the system to attain equilibrium in all the cases reported here .
then we took the average values on 15000 steps after the equilibration . to be computational feasible
, we choose the case where @xmath33 , @xmath71 and @xmath72 . in this section ,
we compare the performance of players in nmg with that in mg .
since the system exhibits peculiar properties in nmg with @xmath73 , we delay the discussion about this case to section [ mg0 ] .
let us begin by investigating the properties of the mean attendance as a function of the control parameter @xmath60 in nmg .
obviously , the mean attendance in nmg is similar to mg for all values of @xmath74 .
that is to say , it always fluctuates around @xmath75 such that players can maximize their global profit .
next we evaluate the performance of the players in nmg by studying the variance of the attendance per player @xmath76 versus the control parameter @xmath60 as shown in fig .
[ fig : f2 ] .
( note that the variance of the two choices must be the same by symmetry for a 2-choice game ; and all the variance mentioned in this section are divided by the number of player @xmath0 for objective comparison . )
we find that the variance in nmg is always much smaller than that in mg for small @xmath60 no matter what is the value of @xmath51 . in other words , players perform much better in nmg than in mg through the introduction of their own local information when @xmath60 is small . when the reduced strategy space size is relatively small comparing with the number of strategies at play , the fluctuation of the attendance is large in mg due to the overcrowding effect of player s strategies @xcite .
under the overcrowding effect , players tend to choose the same alternative for the same global history since they use some very similar strategies . however , in nmg , players can choose different alternatives even they use the same strategy since they consider both their local information and the global information in deciding their choice .
such phenomenon will dilute the overcrowding effects of the strategies in nmg and so players can perform much better than in mg when @xmath60 is small .
when the control parameter @xmath60 increases , the overcrowding effect will be suppressed in both nmg and mg . in mg , the maximal cooperation amongst the players can be achieved subsequently when the number of strategies at play is approximately equal to the reduced strategy space size . however , in nmg , players need to cooperate with each other through both the global information and their local information .
these mixed types of information obtained from different sources makes the cooperation amongst all players becomes much more difficult for all value of @xmath51 .
so the variance in nmg is larger than in mg around the critical point @xmath77 and it tends to the coin toss value ( which is the variance resulting fromplayers making random choices throughout the game ) when @xmath60 increases further .
[ fig : f2 ] shows the variance of attendance per player @xmath76 as a function of the control parameter @xmath60 in nmg for different ratio of @xmath6 to @xmath79 .
when @xmath80 , the variance becomes larger as @xmath79 increases regardless of the value of the control parameter @xmath60 .
moreover , the variance tends to the coin - toss value for @xmath81 over all range of @xmath60 .
furthermore , when @xmath48 , the performance of players becomes better if more local information is available no matter what is the value of @xmath60 . to account for this phenomenon
, we should consider the structure of strategies in the reduced strategy space @xmath41 .
when @xmath82 , each strategy is composed of more segment strategies belonging to a smaller strategy space @xmath64 as @xmath79 increases . therefore , players are more likely to use similar strategies and the overcrowding effect of strategies will dominate which results in large fluctuations of the attendance .
similarly , when @xmath48 , the segment strategies are drawn from @xmath63 and thus the variance becomes larger as @xmath6 increase due to the overcrowding effect of strategies . in particular , when @xmath81 , each strategy is composed of the largest number of segment strategies drawn from the smallest strategy space ( @xmath63 or @xmath64 ) and so the variance is the largest for such case .
when @xmath73 , only local information is available to each player . from numerical simulation ,
the variance in nmg with @xmath73 is found to be smaller than in mg and nmg with @xmath78 when the control parameter @xmath60 is small and approaches zero .
it implies that the cooperation amongst players in nmg with @xmath73 is much better than in mg and nmg with @xmath78 for small @xmath60 .
in fact , player s cooperation in nmg with no global information is resulted from their local interaction . to investigate the local interaction ,
we calculate the correlation function for the local minority choices of two neighboring players as follows : @xmath83 ^ 2 \rangle - \langle \omega_l^{i}(t ) \rangle^2},\ ] ] where @xmath84 is the local minority choice of the @xmath12th player at time @xmath4 .
we reveal that the correlation function @xmath85 is equal to infinity for many neighboring players when @xmath73 .
that is to say , many players always choose the same alternative as their nearest neighbors throughout the game .
in such case , we found that the local histories of these players remain the same such that their local minority choices are completely frozen .
in fact , the decision of a frozen player is dominated by the components of segment strategies corresponding to @xmath86 or @xmath87 whenever @xmath73 . for instance , suppose the components of segment strategies corresponding to @xmath88 equal @xmath89 for all @xmath31 strategies of two nearest neighbors .
once their local histories are @xmath90 , their own local minority will then be @xmath91 in this turn and their local histories will be @xmath90 again at the next time step .
it is a fixed point of the dynamics leading to the dimension reduction of the reduced strategy space .
if similar configurations arise along the ring , the choices of some players can be determined before each turn because their actions are frozen .
the freezing effect of players will minimize the fluctuations of attendance since it is only contributed by few non - frozen players .
indeed , it is the negligence of global information that allows the strong local interactions amongst players who are connected on a ring . on the contrary
, players can only weakly interact with each other locally if they are placed on a dynamically random chain in which each player is connected to two randomly chosen players with all the connections between players change at every time step .
hence , whenever @xmath73 , the freezing effect of players disappears and the variance tends to the coin - toss value for the case of the dynamically random chain as shown in fig .
[ fig : f3 ] .
on the other hand , numerical results show that the probability for the freezing of players decreases exponentially when @xmath6 increases for @xmath48 .
it is because each strategy is composed of more number of segment strategies and the disturbance from the global information becomes more significant when more global information is available .
so the variance becomes larger as @xmath79 decreases for @xmath48 . in order to investigate the crowding effect in nmg
, we compare the numerical results of the variance with the predictions by the crowd - anticrowd theory .
we find a large discrepancy between them whenever @xmath92 or @xmath93 although their trends are consistent with each other for all value of @xmath74 .
according to the crowd - anticrowd theory , the variance of attendance for a given realization of the quenched disorder @xmath94 ( i. e. the initial configuration of the system ) is given by @xcite @xmath95 where @xmath96 denotes the sum over the strategies in the corresponding reduced strategy space , @xmath97 is the action of strategy @xmath98 to the history @xmath99 at time @xmath4 ( @xmath100 corresponds to the two alternatives ) and @xmath101 is the number of players using strategy @xmath98 at time @xmath4 . in mg
, @xmath102 should be equal to zero and thus is dropped .
( for mg , the variance in fact plays the role of the effective hamiltonian of the spin - glass - like system while the strategies can be interpreted as the quenched disorder @xcite ; thus we can state that the four - point correlation function is dropped in the effective hamiltonian of the system for mg . )
such expectation is based on the fact that the number of players using an uncorrelated strategy pair are , on average of the history @xmath99 , independent from the response of the strategy pair to the history in mg . however , from numerical simulation , we find that @xmath103 is equal to a large negative number for nmg with @xmath92 . in other words ,
the uncorrelated segment strategy pairs are no longer independent with each other . through the local interaction
, players tend to use the uncorrelated segment strategy pairs whenever the pairs choose the opposite choice .
it is a favorable solution of the dynamics because the uncorrelated segment strategies can now contribute to the crowd - anticrowd cancellation which will in turn lead to the further reduction of the fluctuation of attendance .
in fact , the crowd - anticrowd cancellation by the uncorrelated strategies is only possible in nmg since players have the freedom to choose which uncorrelated segment strategies to be used on the basis of the local information .
we can correct the semi - analytical results of the crowd - anticrowd theory by counting the contribution of both the anti - correlated and uncorrelated segment strategies to the crowd - anticrowd cancellation .
the corrected semi - analytical results is found to match with the numerical one as shown in fig .
[ fig : f4 ] .
so we conclude that the discrepancy is mainly due to the crowd - anticrowd cancellation by the uncorrelated strategies for @xmath92 . when @xmath93 , the discrepancy of the predictions of the crowd - anticrowd theory from the numerical results is due to the freezing of players .
the ergodicity of the local histories of frozen players is broken down under the freezing effects .
in other words , some possible states of the local histories will never be visited for frozen players since the effective dimension of the strategy space is reduced .
of course , this will violate the ergodicity assumption in the crowd - anticrowd theory .
so the results predicted by the crowd - anticrowd theory show a large discrepancy from the numerical results when @xmath93 .
in summary , we find that nmg exhibits a remarkably different behavior from mg . in nmg with non - zero global information , selfish players on the ring can cooperate with each other to reduce their loss ; moreover , their cooperation is much better than in mg by using the local information that are disseminated through the ring except when the number of strategies at play is approximately equal to the strategy space size .
such phenomenon is believed to be due to the dilution of the crowding effect of players which is resulted from their local interactions on the ring . in nmg with no global information , many players on the ring are found to be frozen where their action and also their local minority choice remain the same throughout the game .
we reveal that the freezing occurs when their decisions are dominated by the components of segment strategies corresponding to @xmath104 or @xmath87 .
such domination arises because of the strong local interaction on the ring due to the absence of the global information . on the other hand
, we find that the predictions of the crowd - anticrowd theory deviate very much from the numerical results for nmg . such discrepancy is found to be due to the crowd - anticrowd cancellation contributed by the uncorrelated strategies in nmg whereas it is impossible for the uncorrelated strategies to contribute to the crowd - anticrowd cancellation in mg .
in fact , all the above arguments should still be valid when players are connected on a network with different topology in nmg provided that each player always obtains the local information from the same neighbors ; and thus we believe that the cooperative behavior of players are similar in all these cases .
finally , we would like to point out that the order parameter is also worth to be studied for nmg . from this study
, we may learn whether there is a phase transition from a symmetric phase to an asymmetric phase as the complexity of the system increases just like that in mg .
99 d. challet and y. c. zhang , physica a * 246 * , 407 ( 1997 ) .
y. c. zhang , europhys .
news * 29 * , 51 ( 1998 ) .
w. b. arthur , amer .
papers and proc . *
84 * , 406 ( 1994 ) .
d. challet and y. c. zhang , physica a * 256 * , 514 ( 1998 ) .
r. savit , r. manuca and r. riolo , phys .
lett . * 82 * , 2203 ( 1999 ) .
n. f. johnson , s. jarvis , r. jonson , p. cheung , y. r. kwong and p. m. hui , physica a * 258 * , 230 ( 1998 ) .
m. hart , p. jefferies , n. f. johnson and p. m. hui , physica a * 298 * , 537 ( 2001 ) .
m. hart , p. jefferies , n. f. johnson and p. m. hui , eur .
j. b * 20 * , 547 ( 2001 ) .
m. l. hart and n. f. johnson , cond - mat/0212088 .
m. paczuski , k. e. bassler and a. corral , phys .
lett . * 84 * , 3185 ( 2000 ) .
t. kalinowski , h .- j .
schulz and m. briese , physica a * 277 * , 502 ( 2000 )
. s. moelbert and p. de los rios , physica a * 303 * , 217 ( 2002 ) .
e. burgos , h. ceva and r. p.
j. perazzo , cond - mat/0212635 .
quan , b .- h .
wang , p. m. hui and x .- s .
luo , physica a * 321 * , 300 ( 2003 ) . f. k. chow and h. f. chau , physica a * 319 * , 601 ( 2003 ) .
f. chau and f. k. chow , physica a * 312 * , 277 ( 2002 ) .
f. k. chow and h. f. chau , cond - mat/0210608 .
d. challet and m. marsili , phys .
e * 60 * , r6271 ( 1999 ) .
d. challet , m. marsili and r. zecchina , phys .
lett . * 84 * , 1824 ( 2000 ) . | to study the interplay between global market choice and local peer pressure , we construct a minority - game - like econophysical model . in this so - called networked minority game model ,
every selfish player uses both the historical minority choice of the population and the historical choice of one s neighbors in an unbiased manner to make decision .
results of numerical simulation show that the level of cooperation in the networked minority game differs remarkably from the original minority game as well as the prediction of the crowd - anticrowd theory .
we argue that the deviation from the crowd - anticrowd theory is due to the negligence of the effect of a four point correlation function in the effective hamiltonian of the system . |
consider the capillary problem @xmath5 where @xmath6 is a region in @xmath7 with a corner at @xmath8 , @xmath9 @xmath10 and @xmath11 are constants , @xmath12 is the exterior unit normal on @xmath13 and @xmath14 is a function of position on @xmath15 the surface @xmath16 describes the shape of the static liquid - gas interface in a vertical cylindrical tube of cross - section @xmath6 ; see @xcite for background .
we are interested in the behavior of solutions to ( [ eq : abasic ] ) , ( [ eq : bbasic ] ) in a neighborhood of a corner point of the boundary .
we take the corner point to be @xmath17 let @xmath18 , where @xmath19 is the ball of radius @xmath20 about @xmath8 .
polar coordinates relative to @xmath8 will be denoted by @xmath21 and @xmath22 .
we assume that @xmath23 is piecewise smooth and that @xmath24 consists of two arcs @xmath25 and @xmath26 , whose tangent lines approach the lines @xmath27 and @xmath28 , respectively , as the point @xmath8 is approached .
the points where @xmath29 intersect @xmath23 are labeled @xmath30 and @xmath31 ; also , @xmath32 set @xmath33 let @xmath34 be an arclength parametrization of @xmath25 and @xmath35 be an arclength parametrization of @xmath36 each measured from the corner at @xmath37 so that @xmath38 let @xmath39 be an arclength parametrization of @xmath40 and @xmath41 be an arclength parametrization of @xmath42 each measured from the corner at @xmath43 define @xmath44 for @xmath45 , the corner will be said to be _ convex _ and for @xmath46 , the corner will be said to be _
nonconvex_. in @xcite , the existence of radial limits of a bounded solution @xmath47 to ( [ eq : abasic ] ) that satisfies ( [ eq : bbasic ] ) on the smooth portions of @xmath23 is proven provided that @xmath0 was bounded away from @xmath1 and @xmath2 , and for a convex corner an additional condition is satisfied coupling @xmath48 and @xmath49 . in this paper
, we eliminate the requirement that @xmath0 is bounded away from @xmath1 and @xmath50 an additional condition must still be satisfied at a convex corner .
the radial limits of @xmath47 will be denoted by @xmath51 and @xmath52 , which are the limits of the boundary values of @xmath47 on the two sides of the corner if these exist .
[ one ] let @xmath47 be a bounded solution to ( [ eq : abasic ] ) satisfying ( [ eq : bbasic ] ) on @xmath53 be discontinuous at @xmath54 if @xmath55 then @xmath56 exists for all @xmath57 if @xmath58 and there exist constants @xmath59 satisfying @xmath60 so that @xmath61 for all @xmath62 for some @xmath63 , then again @xmath56 exists for all @xmath64 . furthermore , in either case
, @xmath56 is a continuous function on @xmath65 which behaves in one of the following ways : \(i ) there exist @xmath66 and @xmath67 so that @xmath68 and @xmath69 is constant on @xmath70 $ ] and @xmath71 and strictly increasing or strictly decreasing on @xmath72\cap ( -\alpha,\alpha)$ ] .
label these case(i ) and case(d ) , respectively .
\(ii ) there exist @xmath73 so that @xmath74 , and @xmath69 is constant on @xmath70 , [ \alpha_{l } , \alpha_{r}]$ ] , and @xmath71 and either increasing on @xmath75\cap ( -\alpha,\alpha)$ ] and decreasing on @xmath76\cap ( -\alpha,\alpha)$ ] or decreasing on @xmath75\cap ( -\alpha,\alpha)$ ] and increasing on @xmath77\cap ( -\alpha,\alpha)$ ] .
label these case ( i d ) and case ( di ) , respectively . in theorem 1 of @xcite and theorem [ one ]
above , the existence of two intervals @xmath78 $ ] and @xmath79 on which @xmath80 is constant ( i.e. `` side fans '' ) is established but the relationship between the sizes of these side fans and the contact angle is unclear . theorem 2 of @xcite establishes lower bounds on these sizes when the @xmath81 are assumed to exist .
( in @xcite , these lower bounds were shown to be the actual sizes of the side fans . )
what happens if the limits of @xmath0 at @xmath8 do not exist ?
theorem [ two ] and corollary [ cor1 ] provide lower bounds in this situation . for @xmath82
define @xmath83 notice that @xmath84 [ two ] let @xmath47 be a bounded solution to ( [ eq : abasic ] ) satisfying ( [ eq : bbasic ] ) on @xmath53 which is discontinuous at @xmath54 assume @xmath56 exists for all @xmath57 then @xmath56 is a continuous function on @xmath65 which behaves as described in * ( i ) * or * ( ii ) * of theorem [ one ] , there exist fans of constant radial limits adjacent to each tangent direction at @xmath8 and , in terms of the cases labeled in theorem 1 , the sizes of the side fans @xmath85 and @xmath86 satisfy the following conditions : * @xmath87 for all @xmath88 for ( i ) and ( di ) .
* @xmath89 for all @xmath90 for ( d ) and ( di ) .
* @xmath91 for all @xmath90 for ( i ) and ( i d ) . *
@xmath92 for all @xmath88 for ( d ) and ( i d ) .
since @xmath47 is bounded and the prescribed mean curvature is @xmath93 there exist @xmath94 and @xmath95 such that @xmath96 when @xmath97 theorem [ one ] is a consequence of @xcite .
suppose now that @xmath98 as in 2.1 of @xcite , we set @xmath99 and define @xmath100 and @xmath101 as follows : @xmath102 @xmath103 let @xmath104 , v\in \left[\frac{\pi}{2 } , \frac{3\pi}{2}\right]\right\rbrace\ ] ] be the inner half of a torus with axis of symmetry @xmath105 , major radius @xmath106 , and minor radius @xmath107 .
the mean curvature of @xmath108 at the point @xmath109 with respect to the exterior normal is @xmath110 and so we have @xmath111 now set @xmath112 and @xmath113 .
let @xmath114 be functions whose graphs satisfy @xmath115 and @xmath116
. then , from ( [ r1 ] ) , we have @xmath117 let @xmath118 be given by @xmath119 which is the rotation about @xmath120 through the angle @xmath121 .
now for each @xmath122 $ ] , let @xmath123 be given by @xmath124 which is the translation taking the point @xmath125 to the @xmath120 .
now define @xmath126 .
let @xmath127 .
allow @xmath128 to denote the angle that the upward tangent ray to @xmath129 makes with the negative @xmath130-axis and @xmath131 to denote the angle that the upward tangent ray to @xmath132 makes with the positive @xmath130-axis .
let @xmath133 be functions defined by @xmath134 .
allow @xmath135 to denote the modulus of continuity of @xmath136 .
then @xmath135 is also the modulus of continuity of @xmath137 , and @xmath138 for each @xmath122 $ ] .
let @xmath139 and allow @xmath140 to be the closure of @xmath141 in @xmath142 .
as in 2.2 of @xcite , there exists an isothermal parametrization @xmath143 given by @xmath144 such that @xmath145 , @xmath146 , and ( @xmath147 ) - ( @xmath148 ) of @xcite hold , where @xmath149 by @xmath150 of @xcite , if we let @xmath151 for @xmath152 , then @xmath153 . from ( @xmath154 of @xcite ,
there exists a connected arc @xmath155 that @xmath156 maps strictly monotonically onto @xmath157 let the endpoints of @xmath158 be denoted @xmath159 and @xmath160 there exists points @xmath161 such that @xmath162 @xmath163 @xmath164 maps the arc @xmath165 onto @xmath166 and @xmath164 maps the arc @xmath167 onto @xmath168 we must consider the two cases : a. @xmath169 b. @xmath170 * assume first that ( a ) holds .
* set @xmath171 .
we wish to prove that @xmath172 is uniformly continuous on @xmath173 and hence @xmath172 extends to be continuous on @xmath174 if so , then the existence and behavior of the radial limits of @xmath47 follows as in @xcite .
there are three possible cases : a. @xmath175 and @xmath176 b. @xmath177 and @xmath178 c. @xmath179 or @xmath180 and @xmath181 or @xmath182 * case ( i ) : * let @xmath183 we observe that @xmath184 , @xmath185 , and so @xmath186 . we wish to use the argument in the proof of theorem 2 of @xcite .
since @xmath187 we can choose @xmath188 such that @xmath189 and @xmath190 set @xmath191 and @xmath192 with these choices of @xmath193 and @xmath194 notice that @xmath195 and @xmath196 this implies that for @xmath197 small enough and @xmath198 with @xmath199 we have @xmath200 since @xmath201 there exists @xmath202 such that @xmath203 where @xmath204 is as in 2.1 of @xcite .
for each @xmath205 allow @xmath206 where @xmath207 is the area of @xmath208 let @xmath209 choose @xmath210 such that @xmath211 let @xmath212 from the courant - lebesgue lemma , there exists a @xmath213 such that the arclength @xmath214 of @xmath215 is less than @xmath216 where @xmath217 and @xmath218 set @xmath219 and @xmath220 then , for @xmath221 there exist functions @xmath222 @xmath223 where @xmath224 and @xmath225 . > from (
10 ) of @xcite ( or ( [ r2 ] ) above ) , we have that @xmath226 in @xmath227 and @xmath228 in @xmath229 so in @xmath230 , @xmath231 . on @xmath232 , @xmath233 . as in the proof of theorem 2 of @xcite , @xmath234 where @xmath235 this follows since @xmath233 on @xmath236 by ( 15 ) of @xcite if @xmath237 and no boundary condition on @xmath238 is required if @xmath239 repeat the same argument with @xmath240 and @xmath241 . in the same way as above ,
there exist functions @xmath242 @xmath243 such that @xmath244 for @xmath245 where @xmath246 . then combining ( [ r4 ] ) and ( [ r3 ] ) we get @xmath247 for @xmath248 . as in @xcite
, it follows that @xmath249 is uniformly continuous on @xmath250 * case ( ii ) : * case ( ii ) is simply case ( i ) reflected about the @xmath251-plane and the proof follows as above .
* case ( iii ) : * notice that @xmath252 and so @xmath253 and @xmath254 can not both occur and @xmath255 and @xmath256 can not both occur .
the result follows from this , using the arguments in cases 1 & 2 .
in particular , if @xmath257 then we obtain a supersolution @xmath258 as in case ( i ) ( see figure [ figtwo ] ) and if @xmath259 we obtain a supersolution @xmath258 as in case ( ii ) ( see figure [ figthree ] ) ; if @xmath260 we obtain a subsolution @xmath261 as in case ( i ) and if @xmath262 we obtain a subsolution as in case ( ii ) . ] * now assume ( b ) holds .
* let @xmath263 and let @xmath264 be the closure of @xmath31 in @xmath265 .
let @xmath266 be a conformal or anticonformal map taking @xmath267 onto @xmath268 such that the map @xmath269 has a downward orientation ( i.e. the normal @xmath270 to @xmath141 gives a downward orientation ) .
writing @xmath271 and @xmath272 we have @xmath273 and @xmath274 while @xmath275 for @xmath276 $ ]
. then the argument follows from @xcite and the previous argument here , as explained in @xcite .
@xmath277 ] we first note that if @xmath278 with @xmath279 and @xmath280 and @xmath281 both exist , then it follows from @xcite that @xmath56 exists for all @xmath282 $ ] and @xmath56 is a continuous function of @xmath22 on @xmath283 $ ] which behaves as described in * ( i ) * or * ( ii ) * of theorem [ one ] .
the first part of theorem [ two ] follows from this .
suppose @xmath284 is a decreasing sequence with @xmath285 let @xmath286 and set @xmath287 for @xmath288 then @xmath289 is a subset of the unit ball in @xmath290 by the banach - alaoglu theorem , there exist a subsequence @xmath291 of @xmath284 and a function @xmath292 such that @xmath293 converges weak - star to @xmath294 that is , for each @xmath295 @xmath296 let us define @xmath297 ( almost everywhere on @xmath298 ) .
for any @xmath299 by choosing @xmath300 to be the characteristic function of the interval @xmath301 we see that @xmath302 and , by choosing @xmath300 to be the characteristic function of the interval @xmath303 we see that @xmath304 hence @xmath305 and @xmath306 thus @xmath307 choose a sequence @xmath284 with @xmath308 such that @xmath309 as above , there exist a subsequence @xmath291 of @xmath284 and @xmath310 such that @xmath311 converges weak - star to @xmath312 then @xmath313 @xmath314 * case 1 : * suppose case ( i ) or ( di ) of theorem 1 holds and @xmath315 let us assume there exists @xmath88 such that @xmath316 we shall show that this leads to a contradiction . set @xmath317 choose a sequence @xmath284 with @xmath308 such that @xmath318 as before , there exist a subsequence @xmath291 of @xmath284 and @xmath310 such that @xmath311 converges weak - star to @xmath312 then @xmath313 @xmath314 let @xmath319 where @xmath320 if case ( i ) holds and @xmath321 if case ( di ) holds , and @xmath322 set @xmath323 and define @xmath324 by @xmath325 for @xmath326 .
let @xmath327 be defined on @xmath328 by @xmath329 and let @xmath330 denote the outward unit normal to @xmath331 .
then @xmath332 satisfies the capillary problem @xmath333 t f_{k } \cdot \nu_{k } & = & \cos ( \gamma_{k } ) \hspace*{1.7 in } \mbox{on } \,\ , \partial \omega_{k } \backslash \{{\cal o}\ } \ , .
\end{array}\ ] ] since @xmath334 if @xmath335 and @xmath336 if @xmath337 we see ( e.g. @xcite ; also see @xcite ) that @xmath338 converges locally to the generalized solution @xmath339 ( in the sense of miranda @xcite and guisti @xcite ) of the functional @xmath340 where @xmath341 if case ( i ) holds or case ( di ) holds and @xmath342 for all @xmath343 and @xmath344 with @xmath345 and @xmath346 otherwise .
let us now define the sets @xmath347 these sets have a special structure which follows from the fact that @xmath348 minimizes the functional @xmath349 and @xmath350 minimizes the functional @xmath351 in the appropriate sense ( e.g. @xcite ) .
let @xmath352 denote the ( open ) triangular region whose boundary is the triangle with vertices @xmath353 @xmath354 and @xmath355 and set @xmath356 ( see figure [ figone ] ) .
simple trigonometric computations with @xmath357 show that @xmath358 where @xmath359 is the angle @xmath360 this holds for all @xmath361 taking the limit as @xmath362 and noticing that @xmath363 as @xmath364 we see that @xmath365 or @xmath366 this contradicts the fact that @xmath348 ( locally ) minimizes @xmath367 therefore ( [ assumption1 ] ) is false .
this completes case 1 .
[ figone ] [ figone ] , title="fig : " ] * case 2 .
* suppose case ( d ) or ( di ) of theorem 1 holds and @xmath368 let us assume there exists @xmath90 such that @xmath369 set @xmath370 choose a sequence @xmath284 with @xmath308 such that @xmath371 as before , there exist a subsequence @xmath291 of @xmath284 and @xmath310 such that @xmath311 converges weak - star to @xmath312 then @xmath372 @xmath373 let @xmath374 where @xmath375 if case ( d ) holds and @xmath376 if case ( di ) holds , and @xmath322 set @xmath323 and define @xmath324 by @xmath325 for @xmath326 .
let @xmath327 be defined on @xmath328 by @xmath329 and let @xmath330 denote the outward unit normal to @xmath331 .
then @xmath332 satisfies the capillary problem @xmath333 t f_{k } \cdot \nu_{k } & = & \cos ( \gamma_{k } ) \hspace*{1.7 in } \mbox{on } \,\ , \partial \omega_{k } \backslash \{{\cal o}\ } \ , .
\end{array}\ ] ] since @xmath336 if @xmath377 and @xmath334 if @xmath378 we see that @xmath338 converges locally to the generalized solution @xmath339 ( in the sense of miranda [ 23 ] and guisti [ 13 ] , [ 12 ] ) of the functional @xmath340 where @xmath379 if case ( d ) holds or case ( di ) holds and @xmath342 for all @xmath380 and @xmath344 with @xmath345 and @xmath381 otherwise .
now the set @xmath382 minimizes the functional @xmath383 in the appropriate sense ( e.g. [ 8 ] , [ 12 ] , [ 29 ] ) .
let @xmath352 denote the ( open ) triangular region whose boundary is the triangle with vertices @xmath353 @xmath384 and @xmath355 and set @xmath356 ( see figure [ figtwo ] ) .
simple trigonometric computations with @xmath357 show that @xmath385 where @xmath359 is the angle @xmath360 this holds for all @xmath386 taking the limit as @xmath387 and noticing that @xmath363 as @xmath388 we see that @xmath389 or @xmath390 this contradicts the fact that @xmath348 ( locally ) minimizes @xmath367 therefore ( [ assumption2 ] ) is false .
this completes case 2 . *
* suppose case ( i ) or ( i d ) of theorem 1 holds and @xmath368 let us assume there exists @xmath90 such that @xmath391 set @xmath370 choose a sequence @xmath284 with @xmath308 such that @xmath392 as before , there exist a subsequence @xmath291 of @xmath284 and @xmath310 such that @xmath311 converges weak - star to @xmath312 then @xmath372 @xmath373 let @xmath374 where @xmath375 if case ( i ) holds and @xmath376 if case ( i d ) holds , and @xmath322 set @xmath323 and define @xmath324 by @xmath325 for @xmath326 .
let @xmath327 be defined on @xmath328 by @xmath329 and let @xmath330 denote the outward unit normal to @xmath331 .
then @xmath332 satisfies the capillary problem @xmath333 t f_{k } \cdot \nu_{k } & = & \cos ( \gamma_{k } ) \hspace*{1.7 in } \mbox{on } \,\ , \partial \omega_{k } \backslash \{{\cal o}\ } \ , .
\end{array}\ ] ] since @xmath334 if @xmath377 and @xmath336 if @xmath378 we see that @xmath338 converges locally to the generalized solution @xmath339 ( in the sense of miranda [ 23 ] and guisti [ 13 ] , [ 12 ] ) of the functional @xmath340 where @xmath393 if case ( i ) holds or case ( i d ) holds and @xmath394 for all @xmath380 and @xmath395 with @xmath345 and @xmath381 otherwise .
now the set @xmath396 minimizes the functional @xmath351 in the appropriate sense ( e.g. [ 8 ] , [ 12 ] , [ 29 ] ) .
let @xmath352 denote the ( open ) triangular region whose boundary is the triangle with vertices @xmath353 @xmath384 and @xmath355 and set @xmath397 ( see figure [ figthree ] ) .
simple trigonometric computations with @xmath357 show that@xmath398 where @xmath399 is the angle @xmath360 this holds for all @xmath386 taking the limit as @xmath387 and noticing that @xmath363 as @xmath388 we see that @xmath400 or @xmath401 this contradicts the fact that @xmath350 ( locally ) minimizes @xmath402 therefore ( [ assumption3 ] ) is false .
this completes case 3 .
* suppose case ( d ) or ( i d ) of theorem 1 holds and @xmath315 let us assume there exists @xmath88 such that @xmath403 set @xmath317 choose a sequence @xmath284 with @xmath308 such that @xmath404 as before , there exist a subsequence @xmath291 of @xmath284 and @xmath310 such that @xmath311 converges weak - star to @xmath312 then @xmath405 @xmath314 let @xmath319 where @xmath320 if case ( d ) holds and @xmath321 if case ( i d ) holds , and @xmath322 set @xmath323 and define @xmath324 by @xmath325 for @xmath326 .
let @xmath327 be defined on @xmath328 by @xmath329 and let @xmath330 denote the outward unit normal to @xmath331 .
then @xmath332 satisfies the capillary problem @xmath333 t f_{k } \cdot \nu_{k } & = & \cos ( \gamma_{k } ) \hspace*{1.7 in } \mbox{on } \,\ , \partial \omega_{k } \backslash \{{\cal o}\ } \ , .
\end{array}\ ] ] since @xmath336 if @xmath335 and @xmath334 if @xmath337 we see that @xmath338 converges locally to the generalized solution @xmath339 ( in the sense of miranda [ 23 ] and guisti [ 13 ] , [ 12 ] ) of the functional @xmath340 where @xmath406 if case ( d ) holds or case ( i d ) holds and @xmath407 for all @xmath343 and @xmath395 if case ( i d ) holds and there exists @xmath408 such that @xmath409 now the set @xmath396 minimizes the functional @xmath351 in the appropriate sense ( e.g. [ 8 ] , [ 12 ] , [ 29 ] ) .
let @xmath352 denote the ( open ) triangular region whose boundary is the triangle with vertices @xmath353 @xmath354 and @xmath355 and set @xmath397 ( see figure [ figfour ] ) .
simple trigonometric computations with @xmath357 show that @xmath410 where @xmath399 is the angle @xmath360 this holds for all @xmath386 taking the limit as @xmath362 and noticing that @xmath363 as @xmath364 we see that @xmath411 or @xmath412 this contradicts the fact that @xmath350 ( locally ) minimizes @xmath402 therefore ( [ assumption4 ] ) is false .
this completes case 4 .
the proof of theorem [ two ] is complete .
[ cor1 ] suppose @xmath413;$ ] set @xmath414.$ ] * if @xmath415 and case ( i ) or ( di ) holds , then @xmath416 * if @xmath417 and case ( d ) or ( di ) holds , then @xmath418 * if @xmath419 and case ( i ) or ( i d ) holds , then @xmath420 * if @xmath421 and case ( d ) or ( i d ) holds , then @xmath422 * proof : * ( a ) suppose case ( i ) or ( di ) of theorem [ one ] holds , @xmath423,$ ] @xmath424 and @xmath425 by theorem [ two ] ( a ) , we know that @xmath426 or @xmath427 for all @xmath428 since @xmath429 we may set @xmath430 and obtain @xmath431 which is a contradiction since @xmath432 thus @xmath416 \(b ) suppose case ( d ) or ( di ) of theorem [ one ] holds , @xmath423,$ ] @xmath424 and @xmath433 by theorem [ two ] ( b ) , we know that @xmath434 or @xmath435 for all @xmath436 since @xmath437 we may set @xmath430 and obtain @xmath438 which is a contradiction since @xmath439 thus @xmath418 \(c ) suppose case ( i ) or ( i d ) of theorem [ one ] holds , @xmath423,$ ] @xmath424 and @xmath440 by theorem [ two ] ( c ) , we know that @xmath441 or @xmath442 for all @xmath436 since @xmath443 we may set @xmath444 and obtain @xmath445 which is a contradiction since @xmath446 thus @xmath420 \(d ) suppose case ( d ) or ( i d ) of theorem [ one ] holds , @xmath423,$ ] @xmath424 and @xmath447 by theorem [ two ] ( d ) , we know that @xmath448 or @xmath449 for all @xmath428 since @xmath450 we may set @xmath444 and obtain @xmath451 which is a contradiction since @xmath452 thus @xmath422 the proof of corollary [ cor1 ] is complete .
@xmath277 [ ex1 ] let @xmath453 $ ] and @xmath454 $ ] with @xmath455 and @xmath456 set @xmath457 for each @xmath458 let @xmath459 $ ] and @xmath460.$ ] define @xmath461 so that @xmath0 is defined on @xmath462 by @xmath463 set @xmath464 let @xmath465 be fixed for now .
set @xmath466 ( @xmath467 ) ; notice that @xmath468 then @xmath469 @xmath470 @xmath471 @xmath472 using a similar argument for @xmath473 with @xmath474 @xmath475 we see that @xmath476 [ ex2 ] let @xmath453 $ ] and @xmath454 $ ] with @xmath455 and @xmath456 set @xmath457 for each @xmath458 let @xmath477 @xmath478 and @xmath479 define @xmath480 so that @xmath0 is defined on @xmath462 by @xmath481 then @xmath482 @xmath483 thus @xmath484 and
@xmath485 let @xmath465 be fixed for now .
if we set @xmath486 ( @xmath467 ) , then @xmath487 and so @xmath488 and , if we set @xmath489 ( @xmath467 ) , then @xmath490 and so @xmath491 similar estimates hold on @xmath492 now suppose @xmath493 is any decreasing sequence in @xmath494 converging to zero . for each @xmath475 there
exists a @xmath495 such that @xmath496 and , since @xmath0 is piecewise constant , a direct calculation shows that @xmath497 equals @xmath498 the minumum occurs when @xmath499 and the minimum of @xmath500 is @xmath501 the maximum occurs when @xmath502 and the maximum of @xmath500 is @xmath503 thus @xmath504 and @xmath505 in these examples , we have the same essential limits inferior and superior at @xmath8 and yet @xmath506 and @xmath507 behave differently .
in example [ ex1 ] , we have the `` extreme values '' ( [ bound1 ] ) ; the `` effective '' contact angles in ( a ) and ( b ) of corollary [ cor1 ] are @xmath508 and in ( c ) and ( d ) of corollary [ cor1 ] are @xmath509 on the other hand , in example [ ex2 ] , we have the `` intermediate values '' ( [ bound2 ] ) and ( [ bound3 ] ) .
for example [ ex2 ] , the `` effective '' contact angles in ( a ) and ( b ) of corollary [ cor1 ] are @xmath510 and in ( c ) and ( d ) of corollary [ cor1 ] are @xmath511 where @xmath512 $ ] satisfy @xmath513 if @xmath47 is a bounded solution of ( [ eq : abasic ] ) satisfying ( [ eq : bbasic ] ) on @xmath53 which is discontinuous at @xmath8 and @xmath56 exists for all @xmath514 then bounds on the sizes @xmath515 and @xmath516 of side fans can be computed using corollary [ cor1 ] ; the lower bounds on the sizes of these side fans differ among these three examples .
the last section of @xcite dealt with extensions of ( [ eq : abasic ] ) to equations of prescribed mean curvature .
consider the prescribed mean curvature contact angle problem @xmath517 suppose @xmath518 satisfies ( [ pmc ] ) and ( [ pmc : contact ] ) and also suppose the following conditions hold : 99 a. k. echart & k. e. lancaster , _ on cusp solutions to a prescribed mean curvature equation _
, to appear in the pacific j. math .
a. elcrat and k. e. lancaster , _ boundary behavior of a nonparametric surface of prescribed mean curvature near a reentrant corner _ ,
* 297 * ( 1986 ) , 645650 .
m. entekhabi and k. e. lancaster , _ radial limits of bounded nonparametric pmc surfaces _ , pacific j. math . * vol .
283 * , no . 2 ( 2016 ) , 341351 .
m. entekhabi and k. e. lancaster , _ radial limits of capillary surfaces at corners _ , to appear in the pacific j. math .
r. finn , _ equilibrium capillary surfaces _
, springer - verlag , new york , 1987 .
e. giusti , _ generalized solutions of prescribed mean curvature equations _ , pacific j. math . * 88 * ( 1980 ) , 287321 .
e. guisti , _ minimal surfaces and functions of bounded variation _ , birkhuser , boston , 1984 . k. e. lancaster , _ a proof of the concus - finn conjecture _ , pacific j. math . *
247 * , no . 1 ( 2010 ) , 75108 .
k. e. lancaster , _
remarks on the behavior of nonparametric capillary surfaces at corners _ , pacific j. math . *
258 * , no . 2 ( 2012 ) , 369392 . k. e. lancaster and d. siegel , _ existence and behavior of the radial limits of a bounded capillary surface at a corner _ , pacific j. math . * vol .
176 * , no . 1 ( 1996 ) , 165 - 194 .
correction to figures , pacific j. math .
179 * , no . 2 ( 1997 ) , 397 - 402 . m. miranda , _ superfici minime illimitate _ , ann .
scuola norm .
pisa * 4 * ( 1977 ) , 313322 .
l. simon , _ regularity of capillary surfaces over domains with corners _ , pacific j. math .
* 88 * ( 1980 ) , 363377 .
l. f. tam , _ the behavior of capillary surfaces as gravity tends to zero _ , phd dissertation , stanford university 1984 .
l. f. tam , _ regularity of capillary surfaces over domains with corners : borderline case _ , pacific j. math .
* 124 * ( 1986 ) , 469482 . | the principle existence theorem ( i.e. theorem 1 ) of `` existence and behavior of the radial limits of a bounded capillary surface at a corner '' ( pacific j. math .
* vol .
176 * , no . 1 ( 1996 ) , 165 - 194 ) is extended to the case of a contact angle @xmath0 which is not bounded away from @xmath1 and @xmath2 ( and depends on position in a bounded domain @xmath3 with a convex corner at @xmath4 ) .
the lower bound on the size of `` side fans '' ( i.e. theorem 2 in the above paper ) is extended to case of such contact angles for convex and nonconvex corners . |
in classical differential geometry ; a general helix in the euclidean 3-space , is a curve which makes a constant angle with a fixed direction .
the notion of rectifying curve has been introduced by chen @xcite .
chen showed , under which conditions , the position vector of a unit speed curve lies in its rectifying plane .
he also stated the importance of rectifying curves in physics .
on the other hand , the notion of slant helix was introduced by izuyama and takeuchi @xcite .
they showed , under which conditions , a unit speed curve is a slant helix .
later , ahmet t. ali published a paper in which position vectors of some slant helices were shown @xcite . in @xcite , l. kula ,
et al studied the spherical images under both tangent and binormal indicatrices of slant helices and obtained that the spherical images of a slant helix are spherical helices .
the papers mentioned above led us to study on the notion of rectifying slant helices .
we began with finding the equations of curvature and torsion of a rectifying slant helix .
after that , we constructed a second order linear differential equation to determine position vector of a rectifying slant helix . by solving this equation for some special cases
, we obtained a unit speed family of rectifying slant helices which lie on cones .
the euclidean 3-space @xmath0 is the real vector space @xmath1 with the metric @xmath2 where @xmath3 is a rectangular coordinate system of @xmath0 .
a curve @xmath4 is said to be parametrized by the arclength parameter s , if @xmath5 , where @xmath6 .
then , we call @xmath7 unit speed . consider unit - speed space curve @xmath7 has at least four continuous derivatives , then @xmath7 has a natural frame called frenet frame with the equations below , @xmath8 where @xmath9 is the curvature , @xmath10 is the torsion , and @xmath11 is the frenet frame of the curve @xmath7 .
we denote unit tangent vector field with @xmath12 , unit principal normal vector field with n , and the unit binormal vector field with @xmath13 .
it is possible in general , that @xmath14 for some @xmath15 ; however , we assume that this never happens .
a curve is called a slant helix if its principal normal vector field makes a constant angle with a fixed line in space .
a unit speed curve @xmath7 is a slant helix if and only if the geodesic curvature of the spherical image of the principal normal indicatrix of @xmath7 which is @xmath16 is constant @xcite .
a unit speed curve @xmath7 is called rectifying curve when the position vector of it always lie in its rectifying plane .
so , for a rectifying curve we can write @xmath17 a unit speed curve @xmath7 is congruent to a rectifying curve if and only if @xmath18 for some constants @xmath19 and @xmath20 , with @xmath21 @xcite .
if the position vector of a unit speed slant helix always lies in its rectifying plane we call it a rectifying slant helix . for a rectifying slant helix
we have the following theorem .
let @xmath7 be a unit speed curve in @xmath0 .
then , @xmath22 is a rectifying slant helix if and only if the curvature and torsion of the curve satisfies the equations below ; @xmath23 where @xmath24 , @xmath25 , and @xmath26 let @xmath7 be a unit speed rectifying slant helix in @xmath0 , then the equations in theorem 1 , and theorem 2 exists . if we combine them then we have @xmath27 where @xmath28 is a constant .
so we can write @xmath9 as follows let @xmath7 be a unit speed rectifying slant helix whose principal normal vector field makes a constant angle with a unit vector @xmath33 , then the curvature and torsion of @xmath7 satisfy the equations below ; @xmath34 where @xmath35 let @xmath7 be a unit speed rectifying slant helix then we can write frenet equations as follows , @xmath52 where @xmath53 .
if we divide second equation by @xmath9 we have , @xmath54 by differentiating ( 4 ) , we have @xmath55 by differentiating ( 5 ) and using ( 3 ) we have @xmath56 n=0,\ ] ] with the necessary calculations we easily see @xmath57 so we have ( 6 ) as follows , @xmath58 let us denote @xmath59 .
then ( 7 ) becomes to @xmath60 this completes the proof .
@xmath64 \right],\\ & & n_2(s)=\int\kappa(s ) v_2(s)ds = a_2 \left| c _ { 1 } \right|\sin ( \theta ) \sin\left[\sec ( \theta ) \arctan\left[f(s)\right ] \right],\\ & & n_3(s)=\cos ( \theta ) .
\end{matrix } \end{matrix}\ ] ] on the other hand , let @xmath7 be a unit speed rectifying slant helix , whose principal normal vector field makes a constant angle @xmath65 with @xmath66 .
then , for its principal normal we can write @xmath67 while @xmath63 is a unit vector , @xmath68 .
so , @xmath69
. therefore @xmath70 can be in the form , @xmath74 ds \right)ds,\\ & & \alpha_2(s)=\sin ( \theta)\int \left(\int \kappa(s ) \sin\left[\sec ( \theta ) \arctan\left(c _ 1 s+c_2\right ) \right]ds \right)ds,\\ & & \alpha_3(s)=\int \left(\int \kappa(s)\cos(\theta ) ds \right ) ds , \end{matrix } \end{matrix}\ ] ] @xmath76 } , \\ \alpha _ { 2 } ( s)=-\frac { \cos { ( \theta ) } } { c _ { 1 } } \sqrt { 1+\left ( c _ { 1 } s+c _ { 2 } \right ) ^ { 2 } } \sin { \left [ \sec { ( \theta ) } \arctan { \left ( c _ { 1 } s+c _ { 2 } \right ) } \right ] } , \\ \alpha _ { 3 } ( s)=\frac { 1 } { c _ { 1 } } \sqrt { 1+\left ( c _ { 1 } s+c _ { 2 } \right ) ^ { 2 } } \sin { ( \theta ) . }
\end{matrix } \end{matrix}\ ] ] let @xmath77 be a space curve with the equation below , @xmath78,\\ & \cos(\theta ) \sin \left[\sec { ( \theta ) } \arctan { \left ( c _ { 1 } s+c _ { 2 } \right ) } \right],\\ & -\sin(\theta ) ) , \end{split}\ ] ] where @xmath79 , and @xmath80 then , @xmath22 is a unit speed rectifying slant helix which lies on the cone @xmath81 with direct calculations we have @xmath82 , @xmath83 , and the curvature functions of @xmath7 as , @xmath84 with , @xmath85 and @xmath86 so , @xmath7 is a unit speed spacelike rectifying slant helix .
we also have @xmath87 then , @xmath7 lies on the cone above .
ali t. ahmad , position vectors of slant helices in euclidean space @xmath0 , journal of the egyptian mathematical society volume 20 , issue 1 , pages 1 - 6 , april 2012 .
chen , b. y. , when does the position vector of a space curve always lie in its rectifying plane ?
monthly 110 , 147 - 152 , 2003 .
chen , b. y. , dillen , f. rectifying curves as centrodes and extremal curves , bull .
inst . math .
academia sinica 33 , no . 2 , 77 - 90 , 2005 .
izumiya s. , takeuchi n. , new special curves and developable surfaces , turk .
28 , 153 - 163 , 2004 .
s. izumiya , n. takeuchi , generic properties of helices and bertrand curves , j. geom .
74 , 97 - 109 , 2002 .
kula , l. and yayli , y. , on slant helix and its spherical indicatrix , applied mathematics and computation , 169 , 600 - 607 , 2005 .
kula , l. ; ekmekci , n. yayl , y. and ilarslan , k. , characterizations of slant helices in euclidean 3-space , turkish j. math .
2 , 261273 , 2010 . oneill b. , elementary differential geometry , academic press , 2006 .
struik d. j. , lectures on classical differential geometry , dover , 1961 . | in this paper , we study the position vector of rectifying slant helices in @xmath0 .
first , we have found the general equations of the curvature and the torsion of rectifying slant helices .
after that , we have constructed a second order linear differential equation and by solving the equation , we have obtained a family of rectifying slant helices which lie on cones . *
key words : * rectifying curve , curvature , torsion , slant helix , cone . *
2010 ams subject classification :* 53a04 , 53a05 |
the low - energy sector of qcd has been successfully described over recent years within the chiral lagrangian framework .
chiral perturbation theory ( chpt ) is based on the spontaneously breaking of chiral symmetry @xmath12 with @xmath13 light flavors and provides a consistent , systematic and model - independent scheme to calculate low - energy observables @xcite .
the effective chpt lagrangian is constructed as an expansion of the form @xmath14 where @xmath15 denotes a meson energy scale compared to the chiral scale @xmath16 1 gev . for @xmath17 case ,
the vector group symmetry is broken by the strange - light quark mass difference @xmath18 , although @xmath19 can still be considered as a perturbation compared to @xmath20 , leading to @xmath4 chpt , which reduces formally to @xmath3 in the @xmath21 limit @xcite .
the formalism can also be extended to finite temperature @xmath0 , in order to describe meson gases and their evolution towards chiral symmetry restoration for @xmath0 below the critical temperature @xmath22 @xcite , where @xmath23 180 - 200 mev from lattice simulations @xcite .
the use of chpt in this context is important in order to provide model - independent results for the evolution of the different observables with @xmath0 , supporting the original predictions for chiral restoration @xcite , also confirmed by lattice simulations , which are consistent with a crossover - like transition for @xmath17 ( 2 + 1 flavors in the physical case ) , which becomes of second order for @xmath24 , in the @xmath25 universality class , and first order in the degenerate case of three equal flavors .
the @xmath26 vector group is the isospin symmetry , which is a very good approximation to nature .
however , there are several examples where isospin breaking corrections are phenomenologically relevant , such as sum rules for quark condensates @xcite , meson masses @xcite or pion scattering @xcite .
for a recent review see @xcite .
the two possible sources of isospin breaking are the qcd @xmath27 light quark mass difference and electromagnetic interactions .
both can be accommodated within the chpt framework .
the expected corrections from the first source are of order @xmath28 and are encoded in the quark mass matrix , generating also a @xmath11 mixing term in the @xmath4 lagrangian @xcite .
the electromagnetic interactions are included in the chpt effective lagrangian via the external source method and give rise to new terms @xcite of order @xmath29 , @xmath30 and so on , with @xmath31 the electric charge .
these terms are easily incorporated in the chpt power counting scheme by considering formally @xmath32 , with @xmath33 the pion decay constant in the chiral limit .
the purpose of this paper is to study within chpt isospin breaking effects related to the thermodynamics of the meson gas .
we will be particularly interested in the physical quantities directly related to spontaneous chiral symmetry breaking and its restoration , namely , the quark condensates and their corresponding susceptibilities at finite temperature .
the quark condensate is the order parameter of chiral restoration , but since the transition is a smooth crossover for the physical case , different observables can yield different transition temperatures .
thus , the susceptibilities , defined as derivatives of the condensates with respect to the quark masses , provide also direct information about the transition and its nature , since they tend to peak around the transition point reflecting the growth of correlations .
let us mention some of the motivations we have in mind for the present analysis .
for the physical values of quark and meson masses , we are interested in the effect of the isospin - breaking terms in the light quark condensate @xmath34 and therefore on the chpt estimates of the critical temperature . in addition , in the isospin asymmetric case , one has @xmath35 and in fact @xmath6 can be considered an order parameter for isospin breaking .
actually , isospin is not spontaneously broken in qcd @xcite which means that this order parameter should vanish for @xmath36 and @xmath37 .
this is an important difference with the scalar condensate @xmath34 , which is nonzero in the chiral limit .
it is relevant to estimate the thermal evolution of @xmath6 , since in principle the two condensates melt at different critical temperatures .
a further motivation is the analysis of the three independent susceptibilities , directly related to the isosinglet , connected ( isotriplet ) and disconnected susceptibilities @xcite often discussed in lattice analysis @xcite .
including properly the @xmath38 dependence of condensates is then essential to analyze the temperature and mass evolution of the connected and disconnected pieces measured in the lattice .
in particular , the linear @xmath27 corrections to condensates survive the @xmath36 limit in the susceptibilities .
the contributions coming from @xmath11 mixing in the @xmath4 case belong to this type and are particularly important regarding the temperature dependence .
this is not only interesting for physical masses but also to explore the scaling near the chiral limit , which in lattice studies has been used to investigate the nature of the transition @xcite . in the lattice
works , this scaling may be contaminated by lattice artifacts such as taste breaking in the staggered fermion formalism , which can generate contributions to susceptibilities masking the true scaling behaviour @xcite .
our study provides then a model - independent setup for disentangling these effects and establishes the expected results in the continuum limit .
we will work in chpt to one loop , considering on the same footing the two sources of isospin - breaking . in a previous work @xcite
we have studied the quark condensates at @xmath39 and several related phenomenological aspects of the isospin asymmetric case .
we will refer to that work for more details about the formalism , the numerical values of the low - energy constants and other related issues .
the paper is organized as follows . in section [ sec : form ] we will review the main aspects of the isospin - breaking chpt formalism related to the present work .
our results for the quark condensates at finite @xmath0 both in the @xmath3 and @xmath4 cases are given and analyzed in section [ sec : cond ] .
in that section we explore the temperature dependence of isospin breaking , as well as that of the sum rule relating condensate ratios .
section [ sec : susc ] is devoted to the analysis of the different isospin - breaking scalar susceptibilities and their relation to the connected and disconnected ones . in subsection
[ sec : sumrule ] we provide an interesting sum rule relating the electromagnetic differences in the condensates with the total susceptibility .
we explore the possibility of using that sum rule to estimate the errors in the staggered fermion lattice analysis of the condensates , in connection with the taste breaking effect . in subsection
[ sec : sustempmass ] we make a thorough study of the connected and disconnected contributions to the susceptibility and their dependence with temperature , the quark mass and the isospin ratio @xmath40 .
we pay special attention to the connection of our results with different lattice analysis in the literature .
the effective chiral lagrangian up to fourth order in @xmath15 ( a meson mass , momentum , temperature or derivative ) including electromagnetic interactions proportional to @xmath41 is given schematically by @xmath42 .
the most general second order lagrangian is the familiar non - linear sigma model , including the gauge coupling of mesons to the electromagnetic field through the covariant derivative , plus an additional term proportional to a low - energy constant @xmath43 compatible with the @xmath2 symmetries of the qcd lagrangian @xcite : @xmath44+c{\mbox{tr}}\left[ququ^\dagger\right ] .
\label{l2}\ ] ] here , @xmath45\in su(n_f)$ ] , with @xmath46 the goldstone boson ( gb ) matrix field for pions ( @xmath24 ) plus kaons and @xmath47 ( @xmath17 ) , the latter being the octet member with @xmath48 .
the covariant derivative is @xmath49 $ ] with @xmath50 the em field . @xmath51 and @xmath52 are the quark mass and charge matrices , i.e. , in @xmath4 @xmath53 and @xmath54 . both the mass term and the charge one proportional to @xmath43 in ( [ l2 ] ) break explicitly the chiral symmetry @xmath55 under which @xmath56 with @xmath57 .
the vector symmetry @xmath58 is also broken for unequal quark masses and charges .
thus , in the light sector ( @xmath59 ) the part of the mass term proportional to @xmath60 , the average light quark mass , is also proportional to the identity flavor matrix and therefore invariant under @xmath26 , while the part proportional to the mass difference @xmath61 and @xmath62 , the third isospin generator , is the one carrying out the qcd isospin breaking .
the only remaining symmetry of the lagrangian ( [ l2 ] ) is the @xmath63 @xmath64 corresponding to charge conservation .
working out the kinetic terms in ( [ l2 ] ) allows to relate the low - energy parameters @xmath65 to the leading - order tree level values for the decay constants and masses of the pseudo - goldstone bosons .
for @xmath3 the masses read : @xmath66 in the @xmath4 case , the mass term in ( [ l2 ] ) induces a mixing between the @xmath67 and the @xmath47 fields given by @xmath68 .
this mixing between the two states with @xmath48 will play an important role in what follows .
the kinetic term has then to be brought to the canonical form before identifying the gb masses , which can be easily done by the field rotation @xcite : @xmath69 where the mixing angle is given by : @xmath70 once the above @xmath11 rotation is performed , the @xmath4 tree level meson masses to leading order read : @xmath71,\nonumber\\ m_{k^+}^2&=&m_{k-}^2=(m_s+m_u ) b_0 + 2 c \frac{e^2}{f^2},\nonumber\\ m_{k^0}^2&=&(m_s+m_d ) b_0,\nonumber\\ m_\eta^2&=&2b_0\left[\frac{1}{3}(\hat m + 2m_s)+\frac{2}{3}(m_s-\hat m)\frac{\sin^2\varepsilon}{\cos 2\varepsilon}\right ] .
\label{treemassessu3}\end{aligned}\ ] ] for pions , the main effect in the @xmath72 mass difference comes from the em contribution @xcite , while in the kaon and eta cases the violations of dashen s theorem @xmath73 @xcite ( @xmath36 limit ) indicate that @xmath38 corrections are relevant and must be kept on the same footing as the em ones @xcite .
we emphasize that all the previous expressions hold for tree level lo masses @xmath74 with @xmath75 , in terms of which we will express all our results .
they coincide with the physical masses to leading order in chpt , i.e. , @xmath76 and so on for the meson decay constants @xmath77 .
the fourth - order lagrangian consists of all possible terms compatible with the qcd symmetries to that order , including the em ones .
the @xmath78 lagrangian is given in @xcite for the @xmath3 case , @xmath79 ( contact terms ) and @xmath80 denoting the dimensionless low - energy constants ( lec ) multiplying each independent term , and in @xcite for @xmath4 the lec named @xmath81 and @xmath82 .
the electromagnetic @xmath30 and @xmath83 for @xmath3 are given in @xcite , @xmath84 denoting the corresponding lec , and in @xcite for @xmath4 with the @xmath85 lec .
the relevant terms needed for this work are given in @xcite .
the lec are renormalized in such a way that they absorb all the one - loop ultraviolet divergences coming from @xmath86 and @xmath29 , according to the chpt counting , rendering the observables finite and scale - independent .
the numerical values of the lec at a given scale can be fitted to meson experimental data , except the contact @xmath87 and @xmath88 .
the latter are needed for renormalization but can not be directly measured , reflecting an ambiguity in the observables depending on them .
the origin of this ambiguity is in the very same definition of the condensates in perturbation theory @xcite .
it is therefore convenient to define suitable combinations which are independent of those constants and therefore can be determined numerically .
we will bear this in mind throughout this work and we will try to provide such combinations when isospin - breaking is included .
the numerical values we will use for masses and low - energy constants in the @xmath4 case are the same as in @xcite unless otherwise stated . in @xmath4
they come from the fits performed in @xcite .
the quark condensates for a given flavor @xmath89 at finite temperature @xmath0 are given by : @xmath90 where @xmath91 , @xmath92 is the system volume , @xmath93 the partition function and @xmath94 denotes a thermal average
. we will denote by @xmath95 , the order parameter of chiral symmetry , while @xmath96 behaves as an order parameter of isospin breaking , since it is the expectation value of the part of the mass term in the qcd lagrangian proportional to @xmath38 and @xmath97 , respectively .
it is still invariant under transformations in the third direction of isospin , which reflects electric charge conservation . in chpt to one loop
we obtain then the @xmath3 finite temperature extension of the @xmath39 results in @xcite , which we give also for consistency : @xmath98+\mathcal{o}\left(p^2\right),\nonumber\\ { \langle \bar
q q \rangle}_0&=&-2f^2b_0\left[1-\mu_{\pi^0}-2\mu_{\pi^\pm}+2\frac{m_{\pi^0}^2}{f^2}\left(l_3^r(\mu)+h_1^r(\mu)\right ) + e^2{\cal k}_2^r(\mu)+\mathcal{o}\left(p^4\right)\right ] , \label{condsu2sum } \end{aligned}\ ] ] @xmath99 where : @xmath100,\ ] ] and @xmath101 with @xmath102 . the expression ( [ condsu2sum ] )
contains the leading order tree level term from @xmath103 given by @xmath104 , the one - loop tadpole like contribution @xmath105 , with @xmath106 the free meson thermal propagator , whose finite part yields the combinations @xmath107 ( we follow the same finite-@xmath0 notation as in @xcite ) and the tree level from the fourth order lagrangian , showing up only at @xmath39 , which contains the lec renormalized at the scale @xmath108 of dimensional regularization in the @xmath109 scheme @xcite so that the full expressions for the condensates are finite and scale - independent .
note that @xmath110 includes the contact term @xmath111 .
the isospin breaking in @xmath112 for @xmath3 is purely electromagnetic , showing up explicitly in the @xmath41 terms and implicitly through the pion mass differences .
the temperature dependence is encoded in the functions @xmath113 which increase with @xmath0 and behave near the chiral limit ( @xmath114 ) as @xmath115 $ ] .
note that the effect of the electromagnetic corrections is to decrease the thermal part of @xmath112 , since @xmath116 . on the other hand
, @xmath110 increases for the available estimates of the em lec , reflecting its ferromagnetic nature @xcite .
our first conclusion is then that the critical temperature , estimated as that for which the condensate vanishes , increases with respect to the @xmath37 case , which is also a ferromagnetic - like behaviour induced by the explicit chiral symmetry breaking of the em quark coupling in the qcd action .
a simple estimate of the size of this effect can be obtained by taking the chiral limit @xmath117 so that @xmath118 , @xmath119 and @xmath120 , which gives @xmath121 with the parameters used in @xcite and setting the involved @xmath122 to their maximum expected natural " values @xmath123 .
thus , in principle we expect rather small corrections to chiral restoration from the electromagnetic breaking .
nevertheless , in section [ sec : sumrule ] we will go back to this point in connection with a sum rule relating the charge breaking with the susceptibility , suggesting larger corrections either for higher order transitions or for finite lattice spacing .
the two sources of explicit isospin breaking in the lagrangian show up in the condensate difference ( [ condsu2dif ] ) , which depends linearly on @xmath38 with the contact @xmath124 and vanishes for @xmath36 and @xmath37 in accordance with the absence of spontaneous isospin breaking @xcite mentioned in the introduction . recall that @xmath124 and @xmath125 do not need to be renormalized and are therefore finite and scale - independent .
an important point is that @xmath6 does not receive pion loop corrections in the two - flavor case and it is therefore temperature independent to the one - loop order . in other words ,
isospin breaking in @xmath3 does not change with @xmath0 and the two condensates melt at the same temperature .
this picture will change for @xmath17 due to kaon loops and @xmath11 mixing . in the @xmath4 case
, we calculate to one loop at finite temperature the light and strange condensates , taking into account both @xmath38 and @xmath2 corrections .
the condensates read now : @xmath126+\mathcal{o}\left(p^2\right),\nonumber\\ { \langle \bar q q \rangle}_0^{su(3 ) } & = & -2f^2b_0\left\{1+\frac{8b_0}{f^2}\left[\hat m\left(2l_8^r(\mu)+h_2^r(\mu)\right)+4(2\hat m + m_s)l_6^r(\mu)\right]+e^2\mathcal{k}_{3+}^r ( \mu)\right.\nonumber\\&-&\left.\frac{1}{3}\left(3-\sin^2 \varepsilon\right ) \mu_{\pi^0}-2\mu_{\pi^\pm}-\mu_{k^0}-\mu_{k^\pm}-\frac{1}{3}\left(1+\sin^2 \varepsilon\right)\mu_\eta + \mathcal{o}\left(p^4\right)\right\ } , \label{condsu3sum}\end{aligned}\ ] ] @xmath127+g_1(m_{k^\pm},t)-g_1(m_{k^0},t)\right\}+\mathcal{o}\left(p^2\right ) , \nonumber\\ \langle \bar u u - \bar d d \rangle_0^{su(3)}&=&2f^2b_0\left\{\frac{4b_0}{f^2}(m_d - m_u)\left(2l_8^r(\mu)+h_2^r(\mu)\right)-e^2\mathcal{k}_{3-}^r ( \mu ) \right.\nonumber\\&+&\left .
\frac{\sin 2\varepsilon}{\sqrt{3}}\left[\mu_{\pi^0}-\mu_{\eta}\right]+\mu_{k^\pm}-\mu_{k^0}\right\}+\mathcal{o}\left(p^2\right ) , \label{condsu3dif}\end{aligned}\ ] ] @xmath128+g_1(m_{k^\pm},t)+g_1(m_{k^0},t)\right\}+\mathcal{o}\left(p^2\right),\nonumber\\ \langle \bar s s \rangle_0&=&-f^2b_0\left\{1+\frac{8b_0}{f^2}\left[m_s\left(2l_8^r(\mu)+h_2^r(\mu)\right)+4(2\hat m + m_s)l_6^r(\mu)\right]+e^2\mathcal{k}_{s}^r ( \mu ) \right.\nonumber\\&-&\left . \frac{4}{3}\left
[ \mu_{\pi^0}\sin^2 \varepsilon+\mu_{\eta}\cos^2 \varepsilon\right]-2\left[\mu_{k^\pm}+\mu_{k^0}\right]+\mathcal{o}\left(p^4\right)\right\ } , \label{condsu3str}\end{aligned}\ ] ] where : @xmath129,\nonumber\\ \mathcal{k}_{3-}^r(\mu)&=&\frac{4}{3}\left[k_9^r(\mu)+k_{10}^r(\mu)\right],\nonumber\\ \mathcal{k}_{s}^r(\mu)&=&\frac{8}{9}\left[3\left(k_7+k_8^r(\mu)\right)+k_9^r(\mu)+k_{10}^r(\mu)\right ] .
\label{kscondsu3}\end{aligned}\ ] ] in some of the above terms we have preferred to leave the results in terms of quark instead of meson masses . as in the @xmath3 case ,
the results are finite and scale - independent , which concerns only the @xmath39 part @xcite
. there are some important differences with respect to the @xmath24 case which deserve to be commented .
first , the presence of the @xmath11 mixing angle @xmath130 ( [ mixangle ] ) , as well as the more complicated dependence of meson masses with quark masses ( [ treemassessu3 ] ) , imply that now @xmath38 corrections show up in @xmath131 , apart from the em ones . note also that these corrections in @xmath131 and @xmath132 are at least @xmath133 in the mixing angle , or equivalently in @xmath38 , except for an @xmath134 term in the kaon contribution .
this is so because , apart from the explicit @xmath130 dependence , one has to expand also the meson masses in ( [ treemassessu3 ] ) around @xmath135 .
all the masses depend quadratically on @xmath130 except @xmath136 , @xmath137 with @xmath138 .
since , in addition , @xmath139 for @xmath135 , we end up with the above mentioned term .
another important difference between the two cases is that for @xmath4 there are loop contributions to @xmath140 in ( [ condsu3dif ] ) .
kaon loops arise from the charged - neutral kaon mass difference , while neutral pion and eta ones from @xmath11 mixing . when expanding in @xmath130 now
, the leading order is @xmath141 even for @xmath37 .
these linear terms will be crucial for our analysis of susceptibilities in section [ sec : susc ] .
those loop corrections introduce now a @xmath0 dependence in @xmath140 , unlike the @xmath3 case . as it happened in the @xmath3 case
, we see that @xmath140 in ( [ condsu3dif ] ) vanishes for @xmath41 and @xmath36 , in agreement with @xcite , which we see from our analysis that holds including thermal corrections . at low and moderate temperatures @xmath142 dominates over the kaon and eta contributions in ( [ condsu3dif ] ) , but it should be reminded that @xmath130 in ( [ mixangle ] ) brings up a @xmath143 dependence which reduces the size of the pion term . in order to make a crude estimate ,
let us consider again the chiral limit , but keeping now the leading order in @xmath38 , which we take then very small but nonzero while taking @xmath144 . in this limit
the kaon masses are roughly kept to their physical values , which are well above the critical temperature .
thus , we consider the regime @xmath145 , in which the pion term behaves as @xmath146 .
the kaon and eta contributions go like @xmath147\sqrt{m_{k,\eta}/t}e^{-m_{k,\eta}/t}$ ] @xcite , where we have taken also @xmath37 for simplicity .
the pion term is still dominant due to the exponential suppression of @xmath148,@xmath47 .
however , when compared to the @xmath39 part in that regime , which goes like @xmath149 , we see that the quadratic growth with temperature is controlled by the scale @xmath150 instead of , say , the chiral restoring behaviour of @xmath112 which is controlled by @xmath151 in the chiral limit .
therefore , the order parameter for isospin breaking @xmath140 grows with @xmath0 , although it does so rather softly
. therefore , we do not expect big differences in the melting temperatures of the @xmath152 and @xmath153 condensates .
this is also consistent with the expectation that in the limit where @xmath19 is arbitrarily large , say compared to @xmath154 , the @xmath3 result should be recovered , for which there is no temperature dependence for the condensate difference .
the evolution with temperature of the condensate difference is shown in figure [ fig : uddifsu3 ] for the full case of finite pion mass and both @xmath2 and @xmath1 .
we have used the same set of low - energy constants and parameters as in @xcite , in particular @xmath155 and @xmath156 . for the em lec @xmath157 involved ,
we have displayed in the figure the two curves corresponding to their maximum and minimum expected natural values .
we also show for comparison the result for @xmath36 , which shows that the charge contribution is actually of the same order as the one proportional to @xmath38 .
we see that the @xmath0-dependent amplification of the isospin difference is rather large .
in fact , this order parameter reaches values comparable to its @xmath39 value near the critical temperature , which is about @xmath23 265 mev in @xmath4 chpt .
nevertheless , due to the additional @xmath130 suppressing factor discussed above , this enhancement is not enough to produce a sizable difference in the melting temperature of the @xmath59 condensates , as it is clearly seen in figure [ fig : uddifsu3 ] ( right ) , where we plot the two thermal condensates separately .
the two plots showed in figure [ fig : uddifsu3 ] correspond then respectively to the two order parameters involved here : isospin breaking and chiral restoration . in turn , note that the curves on the right plot are independent of the choice of lec since to this order @xmath158 for @xmath159 .
left : the @xmath160 condensate difference ( isospin breaking order parameter ) at finite temperature in @xmath4 , relative to its @xmath39 value .
right : the two condensates separately.,title="fig : " ] left : the @xmath160 condensate difference ( isospin breaking order parameter ) at finite temperature in @xmath4 , relative to its @xmath39 value .
right : the two condensates separately.,title="fig : " ] the individual condensates in ( [ condsu3sum])-([condsu3str ] ) contain the contact terms @xmath161 .
these terms reflect an ambiguity in the quark condensates , inherent to their renormalization in qcd .
it is therefore very important to deal with combinations of condensates which are free of this ambiguity .
this very same source of ambiguity is also present in lattice simulations at finite @xmath0 . a simple way to get
rid of it is to subtract the @xmath39 contribution .
this is the approach followed by the group @xcite both for condensates and for susceptibilities .
a different possibility is to consider the combination @xmath162 @xcite , or for individual condensates in the isospin breaking case , @xmath163 with @xmath159 .
another sum rule free of contact ambiguities often used in @xmath39 phenomenology to relate condensate ratios @xcite is the following combination : @xmath164= \delta_{sr}(0)+\frac{m_d - m_u}{m_s-\hat m}\frac{1}{f^2}\left[g_1(m_k , t)-g_1(m_\pi , t)\right.\nonumber\\&+&\left.\left(m_k^2-m_\pi^2\right)g_2(m_k , t)\right]-\frac{2ce^2}{f^4}g_2(m_k , t ) , \label{sumrulet}\end{aligned}\ ] ] where @xmath165 , @xmath166 @xmath167 have been neglected and @xmath168 is given in @xcite with both sources of isospin breaking contributing at the same order , not only in the chiral counting but also numerically .
we have seen in section [ sec : cond ] that the @xmath169 ratio receives significant corrections at finite temperature . on the other hand ,
we expect the strange condensate to vary slowly with @xmath0 , from chiral symmetry breaking due to the strange quark mass .
therefore , we expect that the thermal corrections to this sum rule are also sizable . these corrections are plotted in figure [ fig : sr ] for @xmath170 .
they become comparable to the @xmath39 sum rule near the critical temperature . thermal corrections to the sum rule relating condensate ratios . ]
in the isospin - breaking case , the scalar susceptibilities are defined as : @xmath171 so that in the light sector , the relevant one concerning chiral restoration , we have three independent scalar susceptibilities @xmath172 , @xmath173 and @xmath174 . at this point
, it is instructive to recall the definition of the connected and disconnected parts of the susceptibility .
consider the isospin limit with two light identical flavors of mass @xmath175 and @xmath37 .
there is only one light susceptibility in this case , which can be written as : @xmath176 with : @xmath177 where @xmath178 is the dirac operator for every light flavor in the qcd lagrangian and @xmath179 denotes integration over the gluon fields , so that formally @xmath180 where @xmath181 runs over flavor and @xmath182 runs over the space - time , dirac and color indices .
this separation is important for lattice analysis , as we will discuss below , and reflects the contributions with connected and disconnected quark lines , since @xmath183 is the quark propagator .
however , when considering the low - energy representation for the partition function and the susceptibilities in terms of gb fields , it is not so simple to separate the connected and disconnected parts if we take the isospin limit from the very beginning . a possible approach to perform such separation is to work within the partially quenched chpt framework , as discussed in @xcite for the vacuum polarization .
we will however work within the isospin breaking scenario we are considering here , which is very useful for this purpose , as noted first in @xcite for the susceptibilities and used also in @xcite for the vacuum polarization .
the main point is that for @xmath1 : @xmath184 so that one has @xmath185 and since @xmath186 , from ( [ condistot ] ) we have also @xmath187 $ ] .
therefore , with this observation in mind , we define in the isospin - breaking regime the following basis of total , connected and disconnected susceptibilities in terms of the @xmath188 basis in ( [ susdef ] ) : @xmath189 which we can therefore obtain directly from our expressions for the isospin - breaking condensates obtained in the previous section . observe that none of the @xmath157 dependent terms in the condensates depends on the quark masses and therefore the susceptibilities are independent of the em lec . note that according to ( [ susdef ] ) , @xmath190 in ( [ chitot ] ) corresponds to the correlator of the isosinglet condensate @xmath131 , the order parameter of chiral restoration , while the connected contribution @xmath191 is the correlator of the isotriplet @xmath192 , the order parameter for isospin symmetry . a divergence or sudden growth of these susceptibilities would indicate then a phase transition for the corresponding order parameter .
we also remark that the definitions of the connected and disconnected parts in terms of @xmath193 ones are not unique .
we could as well have defined @xmath194 as @xmath195 for arbitrary
@xmath196 , which also reduces to the combination ( [ disdef ] ) in the isospin limit .
we are following the same convention as @xcite .
these formulas can be easily extended to @xmath197 identical flavors , for which @xmath198 . in the following
we will analyze several aspects related to the above defined susceptibilities in different limits . before studying in detail the different susceptibilities , in this subsection we will relate the em corrections ( and actually any charge - like correction to pion masses ) to the condensates , found in section [ sec : cond ] , with the total scalar susceptibility .
consider first the condensate calculated in @xmath3 in ( [ condsu2sum ] ) and let us define the ratio : @xmath199 now note that to one loop , the explicit dependence of the condensate in @xmath41 is only in the @xmath39 part , since the charge dependence in @xmath103 is contained implicitly in the pion mass differences .
therefore , @xmath200 depends on the charge only through the parameter @xmath201 , in which we can further expand ( for the em pion mass difference @xmath202 ) .
taking also into account that the condensate is just the sum of the tadpole contributions for the three pions , we can write : @xmath203+{{\cal o}}(\delta_\pi^2)+{{\cal o}}(p^4)\label{rder}\\&=&-\frac{m_{\pi}^2}{6b_0 ^ 2f^2}\delta_\pi\frac{\partial}{\partial \hat m } \left[{\langle \bar
q q \rangle}_t-{\langle \bar q q \rangle}_0\right]+{{\cal o}}(\delta_\pi^2)+{{\cal o}}(p^4),\end{aligned}\ ] ] which , from the susceptibility definition in ( [ chitot ] ) can be written , to this order , as : @xmath204 .
\label{rsumrulegen}\ ] ] this sum rule relates then pion mass deviations in the condensate with the total scalar susceptibility . note that the above result is written only in terms of the quark mass , the pion mass and decay constant and the charged - neutral mass difference , without specifying if the latter is of electromagnetic origin .
it states that , even though the mass deviation @xmath205 may be small , the corrections to the condensate may be amplified near the phase transition , where the susceptibility is maximum , if such transition is sufficiently strong .
actually , the quantity proportional to @xmath205 on the right hand side of ( [ rsumrulegen ] ) is directly measurable on the lattice @xcite . for the case for the electromagnetic mass difference in @xmath3 discussed in section [ sec : form ] , we have @xmath206 and : @xmath207 with : @xmath208 and @xmath209,\nonumber\\ g_2(m , t)&=&-\frac{dg_1(m , t)}{dm^2}=\frac{1}{4\pi^2}\int_0^\infty dp \frac{1}{e_p } \frac{1}{e^{\beta e_p}-1}. \label{nug2def}\end{aligned}\ ] ] note that @xmath210 is finite , scale - independent and also independent of the @xmath37 lec @xmath211 . in particular
, it is free of the contact - terms ambiguity , which makes it a quantity suitable for physical predictions .
it is also independent of @xmath212 , unlike the individual quark condensates , which have only physical meaning when multiplied by the appropriate quark masses , since the @xmath213 products give meson masses .
in addition , the dependence with the em lec disappears in the difference @xmath200 , which is the quantity directly related to the susceptibility through ( [ rsumrulegen ] ) .
the above relation can also be explored for @xmath4 . however , the connection with the susceptibility is not direct in that case .
the @xmath41 dependence of @xmath214 enters now through @xmath215 and @xmath216 .
however , the condensates ( [ condsu3sum])-([condsu3str ] ) depend on the light quark mass through all the meson masses @xmath217 .
the result is that @xmath200 can be expressed as the susceptibility term in ( [ rsumrulegen ] ) plus a linear combination of @xmath218 and @xmath219 , to this chiral order and neglecting @xmath166 and @xmath220 isospin - breaking corrections in the right hand side .
since the strange quark condensate has a much weaker dependence on temperature than the light one ( or equivalently , we can approximately neglect the thermal functions evaluated on kaon and eta masses ) we expect the @xmath0 behaviour of @xmath200 to be dominated by the light scalar susceptibility also in the @xmath4 case and therefore the sum rule ( [ rsumrulegen ] ) should hold approximately . in this case
we have to one loop : @xmath221+{{\cal o}}(e^4),\nonumber\\ r(0)^{su(3)}&=&1+e^2 \mathcal{k}_{3+}^r ( \mu)-\frac{2c e^2}{f^4}\left[2\nu_{\pi^\pm}+\nu_{k^\pm}\right]+{{\cal o}}(e^4 ) , \label{rsu3}\end{aligned}\ ] ] where the expansion in @xmath41 to leading order allows to express the result in terms of the @xmath222 and @xmath223 masses . as in @xmath3
, @xmath210 is finite , scale - independent and independent of the @xmath37 lec , so that it is free of contact ambiguities .
we compare the above expression with the susceptibility in the @xmath4 case .
as it will become clear in section [ sec : sustempmass ] , the corrections to the total susceptibility @xmath190 from both sources of isospin breaking are small .
actually , we will see that the nlo correction in the qcd breaking is @xmath165 .
the one - loop result is : @xmath224=2\left[3g_2(m_\pi , t)+g_2(m_k , t)+\frac{1}{9}g_2(m_\eta , t)\right ] + { { \cal o}}\left(p^2\right)+{{\cal o}}(e^2)+{{\cal o}}\left(\frac{m_d - m_u}{m_s}\right)^2,\nonumber\\\label{sustotsu3t}\\ \frac{\chi ( 0)}{b_0 ^ 2}&=&\frac{4\hat m^2}{m_\pi^4}\chi ( 0)=16\left[8l_6^r(\mu)+2l_8^r(\mu)+h_2^r(\mu)\right]-4\left[3\nu_\pi+\nu_k+\frac{1}{9}\nu_\eta\right]+{{\cal o}}\left(p^2\right)+{{\cal o}}(e^2)+{{\cal o}}\left(\frac{m_d - m_u}{m_s}\right)^2 \nonumber .\\\label{sustotsu3zero}\end{aligned}\ ] ] the two quantities are compared in figure [ fig : r ]
. the deviations between them are negligible for the range of relevant temperatures .
therefore , although for physical masses the electromagnetic corrections are relatively small , they grow with the susceptibility , which is a model - independent prediction . for comparison ,
taking the value of @xmath225 $ ] from the lattice simulations in @xcite for 2 + 1 flavors with the lattice @xmath22 value gives @xmath226 , not far from the higher temperature values in figure [ fig : r ] , although the chpt curve can not reproduce the susceptibility peak , only the low and moderate @xmath0 behaviour .
these small em corrections for the condensate are in accordance with our simple estimates made in section [ sec : cond ] and translate into a few mev difference in the determination of the critical temperature from the order parameter .
the @xmath210 function in @xmath4 encoding the em corrections of the quark condensate .
we compare it with the normalized light susceptibility for physical quark and meson masses . for these values , @xmath227 . ]
the sum rule ( [ rsumrulegen ] ) has another interesting consequence , regarding lattice simulations . in the staggered fermion lattice formalism , the need to introduce four different copies ( tastes ) for every quark flavor leads to the so called taste violation @xcite .
this is a lattice artifact which in some aspects is similar to the isospin or flavor violations we are analyzing here .
the new tastes enlarge the chiral symmetry group to @xmath228 , producing then 15 pseudo - goldstone bosons plus one massive state ( @xmath229-like ) for every quark flavor .
all these new meson states become degenerate in the continuum limit , where taking the fourth root of the dirac fermion determinant is enough to remove all the spurious copies .
however , for finite lattice spacing @xmath230 , the tree - level masses of those states receive @xmath231 contributions , which break explicitly the chiral group in the lagrangian , only one goldstone boson remaining massless in the chiral limit leaving then a residual @xmath232 or @xmath63 symmetry .
the mechanism is similar to the electric charge one we are analyzing here , by which the charged states receive @xmath233 corrections and the @xmath63 em symmetry remains .
in fact , for the staggered case one can construct a generalized chiral lagrangian including all possible terms compatible with the new symmetry .
this is called staggered chiral perturbation theory @xcite . among the new terms one recognizes contributions of the form @xmath234
$ ] with @xmath235 a given combination of @xmath236 generators , i.e. , like the charge term in ( [ l2 ] ) in @xmath4 .
obviously , the staggered case includes other operator structure and the spectrum of states is more complicated .
however , we can use the sum rule ( [ rsumrulegen ] ) to estimate roughly the expected differences between the lattice staggered condensate and the continuum one , considering the lightest states . for fine enough lattices , one
has @xmath237 for the lightest tastes of squared mass @xmath238 @xcite , where from the two smallest lattices in @xcite we get @xmath239 @xmath240 . with this @xmath241
we can then use ( [ rsumrulegen ] ) to estimate @xmath242 .
consequently , we expect the larger errors coming from this taste violation effect to appear near @xmath22 .
that is indeed the case when we compare lattices of decreasing temporal extent @xmath243 for the condensate data given in @xcite .
more quantitatively , taking also the susceptibility values of @xcite , we get @xmath244 . estimating the @xmath39 part using ( [ r0su2 ] ) with @xmath245
, we get a relative correction for the condensates near @xmath22 with respect to the continuum of about 20% for the @xmath246 data in @xcite and about 12% for the @xmath247 ones in @xcite . following the same idea
, we get a relative difference between the @xmath248 and @xmath246 lattices of around 8% near @xmath22 , which is actually in good agreement with lattice data @xcite .
a direct translation into an error for the critical temperature is not easy to obtain .
if we simply extrapolate the one - loop chiral limit expression @xmath249 , writing the l.h.s of ( [ rsumrulegen ] ) in the one - loop equivalent form ( [ rder ] ) , we get a very rough estimate @xmath250 10 mev for @xmath246 with respect to the continuum , although the chiral limit is not always numerically accurate , as we will actually see in the next section . estimating taste violation effects is important , since they are one of the main sources of the discrepancies between different lattice groups for the determination of the critical temperature .
an important effort has been made over recent years to minimize these effects , not only by considering finer lattices , but also by introducing lattice actions where taste symmetry is reduced @xcite .
the behaviour of condensates and susceptibilities with temperature and quark masses is crucial in order to understand the nature of the chiral phase transition when approaching the chiral region @xmath251 .
lattice simulations have addressed the question of how those quantities scale with @xmath252 and @xmath0 until very recently @xcite .
an essential part of this program concerns the scaling of the connected and disconnected parts of the scalar susceptibility . the disconnected piece is given in terms of closed quark lines and is therefore directly related to @xmath131 and expected to be sensitive to chiral restoration .
actually , near the chiral limit , i.e. , the infrared ( ir ) behaviour , it is known to scale as @xmath253 for @xmath39 and @xmath254 at finite temperature @xcite . the infrared contribution is then controlled by the gb loop contributions .
the situation is not so clear for the connected part , since its infrared divergent piece is proportional to @xmath255 , with @xmath256 the number of identical light flavors @xcite and therefore vanishes for @xmath257 .
its ir finite part contributes in the physical case of massive pions and is actually an important difference between qcd and @xmath258 models in which the lattice scaling fits are based @xcite . besides , the connected contribution receives important false " gb - like corrections coming from taste violation @xcite .
it is therefore important for lattice studies to provide the continuum result for the disconnected and connected susceptibilities in the physical case of 2 + 1 flavors and massive pions .
our present chpt one - loop analysis allows to obtain a model - independent prediction for the low temperature and small mass behaviour of the susceptibilities .
the inclusion of isospin breaking effects is crucial .
in fact , it will be useful for the following discussion to note that : @xmath259 , \label{susconmdelta } \end{aligned}\ ] ] with @xmath61 , so that @xmath191 comes only from the condensate difference in ( [ condsu3dif ] ) and its leading order is obtained from the linear terms in @xmath38 .
now , from our considerations in section [ sec : cond ] , if we neglect for the moment the charge corrections and we take the quark mass derivatives in ( [ condsu3sum ] ) we have , to leading order in @xmath260 , @xmath261 , @xmath262 and @xmath263 where the leading terms @xmath190 , @xmath191 and @xmath264 are @xmath265 in the @xmath260 counting and are given in @xmath4 by ( [ sustotsu3t])-([sustotsu3zero ] ) and : @xmath266-\frac{g_1(m_\pi , t)-g_1(m_\eta , t)}{3(m_\eta^2-m_\pi^2 ) } + { { \cal o}}(p^2)+{{\cal o}}\left(\frac{m_d - m_u}{m_s}\right)^2+{{\cal o}}(e^2 ) , \nonumber\\ \label{chidistlo}\\ \frac{\chi_{dis}(0)}{b_0 ^ 2}&=&\frac{1}{9}\left(288 l_6^r(\mu)-\nu_\eta-27\nu_\pi\right)-\frac{2f^2\left(\mu_\pi-\mu_\eta\right)}{3(m_\eta^2-m_\pi^2)}+{{\cal o}}(p^2)+{{\cal o}}\left(\frac{m_d - m_u}{m_s}\right)^2+{{\cal o}}(e^2 ) , \label{chidiszerolo}\\ \frac{\chi_{con}(t)-\chi_{con}(0)}{b_0 ^ 2}&=&g_2(m_k , t)+\frac{2\left[g_1(m_\pi , t)-g_1(m_\eta , t)\right]}{3(m_\eta^2-m_\pi^2 ) } + { { \cal o}}(p^2)+{{\cal o}}\left(\frac{m_d - m_u}{m_s}\right)^2+{{\cal o}}(e^2 ) , \label{chicontlo}\\ \frac{\chi_{con}(0)}{b_0 ^ 2}&=&2\left(4h_2^r(\mu)+8l_8^r(\mu)-\nu_k\right ) + \frac{4f^2\left(\mu_\pi-\mu_\eta\right)}{3(m_\eta^2-m_\pi^2)}+{{\cal o}}(p^2)+{{\cal o}}\left(\frac{m_d - m_u}{m_s}\right)^2+{{\cal o}}(e^2 ) .
\label{chiconzerolo}\end{aligned}\ ] ] several remarks are in order from the previous expressions .
first , we emphasize that all of them are finite and scale - independent , which can be explicitly checked from the scale dependence of the lec @xcite .
we also note that , unlike the disconnected part , the connected susceptibility does not receive contributions from the mass derivative of pion tadpoles , i.e @xmath267 or @xmath268 .
these turn out to be the dominant ones in the chiral limit ( see below ) and this is what we expected from our previous discussion on the infrared behaviour .
thus , we identify the difference @xmath269 in ( [ chiconzerolo ] ) and the corresponding @xmath270 one in ( [ chicontlo ] ) as the contribution of @xmath11 mixing , while the @xmath271 terms come from the expansion of the kaon contribution in the r.h.s . of ( [ condsu3dif ] ) around the isospin limit . from ( [ susconmdelta ] ) ,
the disconnected part receives in addition a contribution from the sum ( [ condsu3sum ] ) and hence it incorporates the @xmath267 , @xmath272 terms .
note also that the @xmath39 part of @xmath191 depends on the contact term @xmath161 , while that dependence cancels in @xmath194 .
this means that only quantities such as @xmath273 can be unambiguously determined , similarly as the quark condensate case ( see our discussion at the end of section [ sec : cond ] ) .
this is a relevant comment for lattice evaluations of this quantity .
in fact , as a consequence of the vanishing pion terms , we see that @xmath273 vanishes formally in the @xmath21 limit , recovering the pure @xmath3 result that we would get from ( [ condsu2dif ] ) , which holds also for the other susceptibilities taking into account the conversion between the @xmath3 and @xmath4 lec @xcite ( see below ) .
regarding the expansion in @xmath274 , it can be seen from ( [ susconmdelta ] ) , the condensate expressions ( [ condsu3sum ] ) , ( [ condsu3dif ] ) and the meson masses and mixing angle dependence on @xmath154 and @xmath260 , that the linear order in @xmath260 cancels both in the connected and the disconnected parts and so we have written in the previous expressions and in the total susceptibility ( [ sustotsu3t])-([sustotsu3zero ] ) .
it is important to remark that the @xmath275 corrections contain tadpole mass derivative " terms @xmath276 in both @xmath191 and @xmath194 .
however , an important difference between them is that those ir dominant terms do not appear to leading order in the connected contribution .
in fact , since @xmath277 , where the subscript @xmath278 indicates just the @xmath36 and @xmath279 pion , the disconnected part receives an @xmath280 proportional to the second derivative of the pion tadpole and hence more ir divergent , and so on for the thermal part @xmath281 .
this contribution is not present in the connected part to that order .
thus , we expect the isospin - breaking corrections to be larger in the disconnected than in the connected susceptibility .
another pertinent comment is that we have been able to obtain the leading order for @xmath191 and @xmath194 only after considering properly all the @xmath1 contributions and then taking the @xmath36 limit .
however , one can be led to misleading results by setting @xmath36 from the very beginning .
for instance , for two equal masses one could think naively that @xmath282 , from the definition ( [ chitot ] ) .
however , from our previous analysis we see that in the isospin limit what we get actually is @xmath283 and @xmath284 with @xmath191 given in ( [ chicontlo])-([chiconzerolo ] ) i.e , not vanishing for @xmath36 and physical @xmath19 , although formally suppressed in the @xmath21 limit .
thus , in the isospin limit taking just one flavor susceptibility and multiplying by four does not give the total scalar susceptibility , which should be obtained instead by considering the derivative of the full sum of @xmath152 and @xmath153 condensates as given by ( [ chitot ] ) .
note that this correction does not affect to condensates , for which the @xmath285 and @xmath286 difference vanishes for @xmath36 and @xmath279 and therefore @xmath287 in the isospin limit .
this comment may be relevant for certain lattice analysis , where working within the one - flavor " equivalent framework is often done , since in that way it is easy to discuss for instance the taste - breaking effect .
this observation could then help to explain the worse @xmath258 scaling properties of the susceptibility with respect to those of the condensate @xcite . in that work ,
the lattice data for the susceptibility scaling function in the 2 + 1 case suffer from a sizable increase as the strange quark mass is decreased relative to the light one .
that increase is not seen in the quark condensate data and could be due partly to the definition used as we have just explained , since the positive term proportional to @xmath191 increases with @xmath143 , from ( [ chicontlo])-([chiconzerolo ] ) .
this is not the only effect that may cause this
wrong " scaling , since , as pointed out in @xcite taste - breaking induces an artificial infrared pion contribution in @xmath191 which is not present in the continuum , as our above expressions show .
what we are pointing out here is that the one - flavor " @xmath172 and @xmath173 are sensitive to isospin - breaking terms even for @xmath36 and in the continuum , unlike considering for instance the total susceptibility @xmath288 in ( [ chitot ] ) for which the @xmath191 term cancels . in addition , as we will see below , @xmath191 is dominated numerically by its @xmath39 part , which would explain why the anomalous scaling is reduced for the subtracted susceptibility .
the charge corrections to susceptibilities in our previous expressions ( [ sustotsu3t])-([chiconzerolo ] ) arise only through the mass of the charged mesons @xmath222 and @xmath223 .
thus , including the charge amounts to replace @xmath289 in ( [ sustotsu3zero ] ) and ( [ chidiszerolo ] ) and @xmath290 in ( [ sustotsu3zero ] ) and ( [ chiconzerolo ] ) and so on for the thermal parts @xmath291 and @xmath292 .
although for physical values of the electric charge and masses , these represent small perturbative corrections , the fact that near the chiral limit the coefficient of the ir - dominant @xmath267 and @xmath268 reduces in @xmath293 for @xmath294 is the reflection in the scalar susceptibility of the behaviour of the condensate in terms of @xmath205 corrections to the masses analyzed in section [ sec : sumrule ] .
thus , when the mass corrections @xmath205 become sizable , as in the staggered lattice formalism , the susceptibility is reduced by that factor , which eventually would imply that the transition peak or maximum is displaced to a higher temperature , consistent with our analysis in the previous section about the increasing of @xmath22 in the condensates .
recall that one can not just expand the pion terms in @xmath41 and then take the chiral limit , since that expansion assumes that the charge part of @xmath215 is small compared to the quark mass one . before continuing
, we also remark that our above results are compatible with the recent observation @xcite that the connected and disconnected susceptibilities can be inferred from the zero momentum limit of the @xmath295 and @xmath296 correlators calculated previously in staggered chpt @xcite .
the motivation of those works is precisely to estimate the contribution of heavy pion - like tastes to the ir part of @xmath191 , which could mask the scaling behaviour .
the continuum limit of the results in @xcite reveals the same @xmath297 loop contributions as in our expressions ( [ chidistlo])-([chiconzerolo ] ) .
we provide the full chpt result , including the lec contribution necessary to guarantee the finiteness and scale - independency of the results , as well as the analysis of the higher order corrections in isospin breaking .
as discussed above , the behaviour of the susceptibilities near the chiral limit ( ir regime ) is very illuminating regarding their approach to chiral restoration within the @xmath25 or @xmath3 pattern in the continuum , i.e. , without taste breaking effects .
let us consider this regime first for @xmath39 and only for the leading order terms in the isospin expansion in ( [ chidiszerolo])-([chicontlo ] ) , i.e. , we set @xmath298 and @xmath279 in the @xmath39 susceptibilities and consider @xmath299 .
we denote by a superscript ir " the nonvanishing terms in that limit : @xmath300-\frac{1}{16\pi^2}\left(1+\log\frac{m_k^2}{\mu^2 } + \frac{2}{3}\log\frac{m_\eta^2}{\mu^2}\right).\label{chiconzeroir}\end{aligned}\ ] ] the ir divergent @xmath301 term in ( [ chidiszeroir ] ) coincides with the one obtained in @xcite , where a cutoff regularization was used . multiplied by 4 , this is also the ir dominant part in the total susceptibility @xmath190 .
in addition to that term , we obtain here the regular part , not ir divergent but not vanishing in the chiral limit , which provides the dependence with the lec and together with the logarithm gives the consistent scale - independent chpt prediction . as for the connected part , again only the ir divergent contribution is given in @xcite , which as commented above turns out to vanish exactly for two light fermions . here
we also give the regular contribution , also scale - independent , which unlike the disconnected part depends on the contact term @xmath302 . as an interesting consistency check
, we can recover the @xmath3 limit from the above expressions , using the conversion between the lec of @xmath3 and @xmath4 @xcite @xmath303 and @xmath304 , so that @xmath305 and @xmath306 , which is the same result that we would have starting directly from the @xmath3 expressions in ( [ condsu2sum ] ) and ( [ condsu2dif ] ) .
let us consider now the dominant ir thermal contribution in the 2 + 1 flavor case , i.e , apart from @xmath299 we also consider temperatures @xmath307 , so that we neglect all the boltzmann exponentials @xmath308 and expand @xmath309 $ ] and @xmath310 @xcite .
thus , we get : @xmath311^{ir } } { b_0 ^ 2}&=&\frac{3t}{16\pi m_\pi},\label{chidistir}\\ \frac{\left[\chi_{con}(t)-\chi_{con}(0)\right]^{ir}}{b_0 ^ 2}&=&\frac{t^2}{18m_\eta^2}.\label{chicontir}\end{aligned}\ ] ] the disconnected part ( [ chidistir ] ) is again the one obtained in @xcite in the ir limit .
it diverges more strongly than the @xmath39 contribution in ( [ chidiszeroir ] ) in this limit , revealing its critical behaviour .
the growth with @xmath0 is linear over the gb mass scale .
recall that , apart from the heavy masses thermal exponentials , we are neglecting also @xmath312 terms in the disconnected part ( [ chidistir ] ) .
the situation is completely different for the connected contribution , which is regular , albeit not vanishing , in the chiral limit .
the quadratically growing term in ( [ chicontir ] ) survives for @xmath313 against neglected @xmath314 and is dominant over @xmath308 .
it vanishes formally as @xmath21 , recovering the @xmath3 limit . for physical masses
though , an specific and model - independent difference between the @xmath24 and @xmath315 cases is the ( soft ) temperature dependence of the connected susceptibility , the scale that controls its growth being @xmath150 instead of the @xmath316 of the connected part .
this is a consequence of @xmath194 measuring the fluctuations of the chiral restoration order parameter , while @xmath191 is related to those of the isospin - breaking one , i.e. , @xmath6 , which as we have seen in section [ sec : cond ] increases moderately .
note that near the chiral limit we could as well have written the @xmath317 term divided by @xmath318 just by changing the multiplying factor .
keeping @xmath150 reminds its @xmath11 origin , as it is clearly seen in the original expressions in ( [ chicontlo])-([chiconzerolo ] ) . in figure
[ fig : susc ] we plot our numerical chpt results for the susceptibilities , including all the isospin - breaking corrections .
the plots in that figure show the difference with respect to the @xmath39 results , which are collected in table [ tab : susczero ] . at @xmath39
we use the same lec values as in @xcite , which are quoted in the table .
remember that the susceptibilities are independent of the em lec and that the disconnected one is independent of contact terms .
the contact lec @xmath302 appearing in the connected contribution is estimated from resonance saturation arguments @xcite .
the normalization used @xmath319 is the same one used in some lattice works @xcite .
.[tab : susczero ] @xmath39 values for the different susceptibilities in chpt . the isospin limit ( il )
values correspond to @xmath320 and @xmath37 . for the third to sixth rows we fix @xmath40 and @xmath19 and
vary the light to heavy quark mass ratio .
the first and second rows correspond to the physical values .
the lec values used are @xmath321 , @xmath322 at the scale @xmath323 mev . [ cols="^,^,^,^,^",options="header " , ] temperature dependence of disconnected , connected and total susceptibilities in three - flavor chpt , for physical quark and meson masses .
the full blue curves show the results with all the isospin - breaking corrections of higher order included , while in the dashed red ones we display the leading order in the isospin limit .
we also show the ir expressions for the connected and disconnected parts , as well as the deviations of @xmath324 from @xmath325 in the isospin limit ( last plot ) .
the @xmath39 results are given in table [ tab : susczero].,title="fig : " ] temperature dependence of disconnected , connected and total susceptibilities in three - flavor chpt , for physical quark and meson masses .
the full blue curves show the results with all the isospin - breaking corrections of higher order included , while in the dashed red ones we display the leading order in the isospin limit .
we also show the ir expressions for the connected and disconnected parts , as well as the deviations of @xmath324 from @xmath325 in the isospin limit ( last plot ) .
the @xmath39 results are given in table [ tab : susczero].,title="fig : " ] temperature dependence of disconnected , connected and total susceptibilities in three - flavor chpt , for physical quark and meson masses .
the full blue curves show the results with all the isospin - breaking corrections of higher order included , while in the dashed red ones we display the leading order in the isospin limit .
we also show the ir expressions for the connected and disconnected parts , as well as the deviations of @xmath324 from @xmath325 in the isospin limit ( last plot ) .
the @xmath39 results are given in table [ tab : susczero].,title="fig : " ] temperature dependence of disconnected , connected and total susceptibilities in three - flavor chpt , for physical quark and meson masses .
the full blue curves show the results with all the isospin - breaking corrections of higher order included , while in the dashed red ones we display the leading order in the isospin limit .
we also show the ir expressions for the connected and disconnected parts , as well as the deviations of @xmath324 from @xmath325 in the isospin limit ( last plot ) .
the @xmath39 results are given in table [ tab : susczero].,title="fig : " ] we see in the plots that the general features explored in our previous analytical discussion are well reproduced .
first , the @xmath233 and @xmath326 terms neglected in ( [ sustotsu3t])-([sustotsu3zero ] ) and ( [ chidistlo])-([chiconzerolo ] ) are numerically small for the relevant temperature range , for physical values of quark and meson masses .
in fact , as anticipated in our previous discussion , we see that those isospin corrections are larger for the disconnected than the for connected part and are also larger with temperature , all due to the appearance of ir terms proportional to the second derivative of the tadpole in @xmath194 .
remember that the leading order in the isospin limit comes actually from the @xmath327 terms in the condensates .
second , we appreciate qualitatively the linear and quadratic growth with temperature of the disconnected and connected parts respectively , as expected from the infrared analysis .
in fact , we see that although the connected term grows faster , its absolute value is much smaller due to the @xmath150 scale compared to the @xmath316 of the disconnected part .
however , it is important to note that the ir limit expressions ( [ chidistir])-([chicontir ] ) are numerically rather far from the exact ones for the physical pion mass . the difference is larger for the disconnected contribution , since , as stated above , in ( [ chidistir ] ) we are neglecting @xmath328 terms , while in ( [ chicontir ] ) the neglected terms are @xmath314 . in fact , this justifies further our present analysis , since we provide the full expressions beyond the chiral limit . also as discussed above , the infrared limit expressions for @xmath39 given in ( [ chidiszeroir])-([chiconzeroir ] )
survive not only the chiral limit but also the @xmath21 one , @xmath194 still diverging but only logarithmically , which for physical masses makes the two susceptibilities numerically comparable .
this is clearly seen in the values given in the first two rows of table [ tab : susczero ] .
actually , for this very same reason , and following our previous discussion , the deviations of @xmath172 from the naive isospin - limit expectation @xmath329 is much more pronounced at @xmath39 than for finite @xmath0 , as it can be seen by comparing the last two columns in table [ tab : susczero ] which give about a 30% relative difference , while the last plot in figure [ fig : susc ] where we compare their thermal differences give only corrections below 10% .
in fact , this is consistent with our previous discussion about the influence of the connected part in the scaling properties observed in the lattice .
if we consider the subtracted susceptibilities as defined in @xcite from the subtracted condensate @xmath330 , we see that the dependence on the lec disappears in the @xmath331 limit .
remember that in this limit all the @xmath39 contribution of the connected part is absorbed in @xmath332 ( see our previous comments ) and therefore considering the subtracted susceptibility is equivalent to switch off the dominant @xmath39 part of the connected susceptibility .
this is indeed observed in the lattice @xcite since the subtracted susceptibility fits better the expected @xmath258 scaling behaviour than the unsubtracted one .
the variation with the quark mass is displayed for @xmath333 in figure [ fig : suscmass ] and for @xmath39 in table [ tab : susczero ] .
it is important to remark that we have chosen to keep fixed the ratio @xmath334 ( same value used in our previous analysis @xcite ) and @xmath19 , while we vary @xmath335 above and below the physical quark mass ratio . in other words ,
when @xmath336 , @xmath337 while @xmath338 remain fixed .
this is meant to be the relevant limit when approaching chiral restoration .
in addition , since we can write @xmath339 , @xmath130 scales in this limit as @xmath340 .
recall that , although the values of the @xmath341 are fitted to low - energy data with physical masses @xcite , those lec are formally independent of the quark masses @xcite . the same applies to the tree level value of @xmath33 we are using . quark mass dependence of the thermal disconnected , connected and total susceptibilities in three - flavor chpt , for fixed @xmath19 and fixed @xmath40 .
we also show the isospin limit ( il ) @xmath320 and @xmath279 in the smaller light mass case .
the @xmath39 results are given in table [ tab : susczero ] .
for comparison , we also display the infrared limit in the @xmath342 case.,title="fig : " ] quark mass dependence of the thermal disconnected , connected and total susceptibilities in three - flavor chpt , for fixed @xmath19 and fixed @xmath40 .
we also show the isospin limit ( il ) @xmath320 and @xmath279 in the smaller light mass case .
the @xmath39 results are given in table [ tab : susczero ] .
for comparison , we also display the infrared limit in the @xmath342 case.,title="fig : " ] quark mass dependence of the thermal disconnected , connected and total susceptibilities in three - flavor chpt , for fixed @xmath19 and fixed @xmath40 .
we also show the isospin limit ( il ) @xmath320 and @xmath279 in the smaller light mass case .
the @xmath39 results are given in table [ tab : susczero ] .
for comparison , we also display the infrared limit in the @xmath342 case.,title="fig : " ] quark mass dependence of the thermal disconnected , connected and total susceptibilities in three - flavor chpt , for fixed @xmath19 and fixed @xmath40 .
we also show the isospin limit ( il ) @xmath320 and @xmath279 in the smaller light mass case .
the @xmath39 results are given in table [ tab : susczero ] .
for comparison , we also display the infrared limit in the @xmath342 case.,title="fig : " ] as we expected from our previous ir analysis in eqns .
( [ chidistir ] ) and ( [ chicontir ] ) , the light quark mass dependence of the thermal disconnected susceptibility is much stronger than the connected one , as seen clearly in fig.[fig : suscmass ] and as long as we take the limit in the order specified above . in terms of chiral restoration , this anticipates a much stronger growth or peak near @xmath22 for the disconnected part . from the same arguments .
the behaviour of the connected part is expected to be softer near the transition , although growing with @xmath317 for low and moderate temperatures .
we also show in the figure the comparison with the infrared limit for the smaller @xmath343 case , where it can be seen that the curves are now closer than for the physical pion mass case in figure [ fig : susc ] .
in addition , the isospin corrections are also more important for the disconnected part , where they actually increase as @xmath344 is decreased , than for the connected one , where the isospin limit and complete curves are almost indistinguishable in the figure .
the same holds for the @xmath39 contributions in table [ tab : susczero ] . according to our previous discussion ,
this behaviour of the isospin corrections arises from the dominant ir terms @xmath345 as compared to @xmath346 .
this effect is weaker for the @xmath39 contributions since the ir leading corrections @xmath347 ( in @xmath191 ) and @xmath348 ( in @xmath194 ) diverge softly , namely , as @xmath349 and @xmath154 respectively .
thus , although the isospin corrections are amplified in the disconnected susceptibility for large temperatures and small masses , they are still perturbatively under control in the chiral limit .
the limit where @xmath343 vanishes not by taking @xmath144 but keeping @xmath154 fixed and taking @xmath21 is where we recover the pure @xmath3 results , as discussed before . in that case , it is the connected part which is more sensitive to the quark mass variation , vanishing for large @xmath150 , while the disconnected one remains invariant .
although this is formally interesting for connecting the @xmath3 and @xmath4 cases , it is not so relevant for studying the critical behaviour .
in this work we have analyzed the relevant observables regarding chiral symmetry restoration , namely quark condensates and scalar susceptibilities , in the presence of isospin breaking .
we have considered on the same footing the qcd ( @xmath40 mass difference ) and electromagnetic corrections , to one loop in chiral perturbation theory , both in the @xmath3 and @xmath4 sectors .
our analysis provides useful and model - independent results regarding several relevant aspects of isospin breaking and chiral restoration , which may be particularly interesting for lattice studies .
the sum @xmath350 , the order parameter for chiral restoration , receives small isospin - breaking corrections for the physical values of masses and electric charge .
these corrections affect only slightly , less than @xmath351 , to the value of the critical temperature , which they increase as a ferromagnetic response .
the difference @xmath140 is the order parameter of isospin breaking .
it is temperature independent in the @xmath3 limit , but when kaons and eta are included , it shows an increasing behaviour , which in the chiral limit is given by @xmath352 .
the deviations with respect to its @xmath39 value become sizable as the temperature is increased , but they are controlled by a larger energy scale @xmath150 than the typical @xmath353 of @xmath350 .
this large growth of isospin breaking does not reflect in the chiral restoration temperatures of @xmath354 and @xmath355 , which remain close to each other , consistently with the idea that chiral restoration is little affected .
we have also evaluated the temperature corrections to the sum rule relating the @xmath356 and @xmath357 ratios , which is useful because it does not involve undetermined contact low - energy constants .
the corrections in this case come directly from the @xmath169 ratio and are therefore rather large for the temperatures of interest .
a very important part of the present work has been the analysis of scalar susceptibilities in the isospin asymmetric scenario .
we have related the different flavor susceptibilities with the total , quark connected and quark disconnected susceptibilities often used in lattice analysis .
electromagnetic corrections to the quark condensate turn out to be directly related by a sum rule to the total susceptibility and then to the growth of fluctuations , which is meant to be maximum near the critical point .
this sum rule is valid for any small deviation of the pion masses , as for instance the one arising in the staggered lattice formalism due to taste breaking effects .
actually , we have made rough estimates of the corrections to condensates expected from this source , comparing lattices of different sizes among them and with the continuum limit .
these estimates are in good agreement with the errors quoted in the lattice works .
the isospin asymmetric calculation allows for a direct extraction of the connected and disconnected susceptibilities , even in the isospin symmetric limit .
the terms in @xmath140 linearly proportional to @xmath38 give contributions to the connected part not vanishing in the isospin limit and which affect for instance the naive extrapolation of a given flavor susceptibility to the total one .
our analysis provides model - independent predictions for the mass , temperature and isospin dependence of those quantities , which should be recovered in lattice analysis as they approach the continuum limit . in accordance with the behaviour of the corresponding order parameters ,
the disconnected susceptibility shows a linear growth at low and moderate temperatures , infrared divergent near the chiral limit as @xmath9 , whereas the connected one is infrared regular but survives the chiral limit as a growing @xmath10 behaviour .
the chiral or infrared limit gives qualitatively the behaviour as the temperature approaches chiral restoration but numerically is not a good approximation for physical pion masses .
the higher order isospin breaking corrections are quadratic in @xmath38 and are enhanced in the chiral limit for the disconnected susceptibility , as long as @xmath19 and @xmath40 remain fixed .
the chpt susceptibilities reproduce the growing @xmath0-dependence at low and moderate temperatures in a model - independent way .
although they do not show the peaks expected near the transition , our small mass analysis allows to infer that the disconnected part should have a more pronounced peak than the connected one , the latter expected to present a rather soft behaviour .
this difference can be interpreted from the different order parameters that fluctuate in each case : the chiral quark condensate for the disconnected piece and the isospin - breaking one in the connected case . in the formal @xmath3 limit @xmath21
the connected contribution becomes temperature independent , like @xmath140 .
our analysis for the susceptibilities is consistent with previous related work in the literature .
we are grateful to w.unger for useful comments . work partially supported by the spanish research contracts fpa2008 - 00592 , fis2008 - 01323 , ucm - bsch gr58/08 910309 and the fpi programme ( bes-2009 - 013672 ) . c. bernard _
[ milc collaboration ] , phys .
d * 71 * , 034504 ( 2005 ) .
y. aoki , g. endrodi , z. fodor , s. d. katz and k. k. szabo , nature * 443 * , 675 ( 2006 ) .
y. aoki , s. borsanyi , s. durr , z. fodor , s. d. katz , s. krieg and k. k. szabo , jhep * 0906 * , 088 ( 2009 ) .
a. v. smilga and j. j. m. verbaarschot , phys .
d * 54 * , 1087 ( 1996 ) . c. e. detar and r. gupta [ hotqcd collaboration ] , pos * lat2007 * , 179 ( 2007 ) . c. detar and u. m. heller , eur .
j. a * 41 * , 405 ( 2009 ) .
s. ejiri _ et al .
_ , phys .
d * 80 * , 094505 ( 2009 ) .
w. unger [ rbc - bielefeld collaboration ] , pos * lat2009 * , 180 ( 2009 ) .
a. g. nicola and r. t. andres , arxiv:1009.2170 [ hep - ph ] .
t. das , g. s. guralnik , v. s. mathur , f. e. low and j. e. young , phys .
lett . *
18 * , 759 ( 1967 ) .
j. bijnens and j. prades , nucl .
b * 490 * , 239 ( 1997 ) .
g. amoros , j. bijnens and p. talavera , nucl .
b * 602 * , 87 ( 2001 ) .
m. della morte and a. juttner , jhep * 1011 * , 154 ( 2010 ) .
a. juttner and m. della morte , pos * lat2009 * , 143 ( 2009 ) . c. aubin and c. bernard , phys .
d * 68 * , 034014 ( 2003 ) .
s. borsanyi , z. fodor , c. hoelbling , s. d. katz , s. krieg , c. ratti and k. k. szabo [ wuppertal - budapest collaboration ] , jhep * 1009 * , 073 ( 2010 ) . | we analyze quark condensates and chiral ( scalar ) susceptibilities including isospin breaking effects at finite temperature @xmath0 . these include @xmath1 contributions as well as electromagnetic ( @xmath2 ) corrections , both treated in a consistent chiral lagrangian framework to leading order in @xmath3 and @xmath4 chiral perturbation theory , so that our predictions are model independent . the chiral restoration temperature extracted from @xmath5
is almost unaffected , while the isospin breaking order parameter @xmath6 grows with @xmath0 for the three - flavor case @xmath4 .
we derive a sum rule relating the condensate ratio @xmath7 with the scalar susceptibility difference @xmath8 , directly measurable on the lattice .
this sum rule is useful also for estimating condensate errors in staggered lattice analysis .
keeping @xmath1 allows to obtain the connected and disconnected contributions to the susceptibility , even in the isospin limit , whose temperature , mass and isospin breaking dependence we analyze in detail .
the disconnected part grows linearly , diverging in the chiral ( infrared ) limit as @xmath9 , while the connected part shows a quadratic behaviour , infrared regular as @xmath10 and coming from @xmath11 mixing terms .
this smooth connected behaviour suggests that isospin breaking correlations are weaker than critical chiral ones near the transition temperature .
we explore some consequences in connection with lattice data and their scaling properties , for which our present analysis for physical masses , i.e. beyond the chiral limit , provides a useful model - independent description for low and moderate temperatures . |
[ section1 ] unlike traditional imaging systems , hyperspectral imaging ( hsi ) sensors @xcite,@xcite acquire a scene with several millions of pixels in up to hundreds of contiguous wavelengths .
such high resolution spatio - spectral hyperspectral data , i.e. , three - dimensional ( 3d ) datacube organized in the spatial and spectral domain , has an extremely large data size and enormous redundancy , which makes compressive sensing ( cs ) @xcite a promising solution for hyperspectral data acquisition . to date , most existing designs for cs - based hyperspectral imagers can be grouped into frame - based acquisition in the spatial direction and pixel - based acquisition in the spectral direction .
while a lot of reconstruction approaches for these two acquisition schemes have been proposed , most existing algorithms can only take advantage of the spatial and spectral information of hyperspectral data from the aspect of sparsity ( or joint - sparsity ) . because the foundation of these algorithms is built on conventional cs , which reconstructs the signals by solving a convex programming and proceeds without exploiting additional information ( aside from sparsity or compressibility ) @xcite .
for hyperspectral data , the spatial and spectral correlations , which not only reflect in the correlation between the sparse structure of the data ( i.e. , structured sparsity ) , but also in the correlation between the amplitudes of the data , can be used to provide helpful prior information in the reconstruction processed and assist on increasing the compression ratios . in this paper
, the structured sparsity and the amplitude correlations are considered jointly by assuming that spatially and spectrally correlated data satisfies simultaneous low - rank and joint - sparse ( l&s ) structure . using a structured l&s factorization
, we propose an iterative approximate message passing ( amp ) algorithm @xcite,@xcite , in order to enable joint reconstruction of the data with the practical compression ratio that is substantially higher than the state - of - the - art .
specifically , in section [ section2 ] , we introduce the structured factorization representation of the l&s model . in section [ section3 ] ,
we propose a novel amp - based approach , called l&s - approximate message passing ( l&s - amp ) , that decouples the global inference problem into two sub - problems .
one sub - problem considers the linear inverse problem of recovering the signal matrix from its compressed measurements .
another sub - problem exploits the l&s structure of the signal matrix . then a recently proposed
turbo amp " framework @xcite is used to enable messages to pass between these two phases efficiently .
section [ section4 ] presents simulation results with real hyperspectral data that support the potential of the approach to considerably reduce the reconstruction error . in section [ section5 ] , we conclude the paper .
[ section2 ] in this section , we first present the problem for compressive hyperspectral imaging . then , we propose a structured l&s factorization model for the signal matrix , which will be later exploited to acquire any hsi with very few measurements , via a novel joint reconstruction approach .
owing to the inherent 3d structure present in the hyperspectral datacube and the two - dimensional nature of optical sensing hardware , cs - based hyperspectral imagers generally capture a group of linear measurements across either the 2d spatial extent of the scene for a spectral band or the spectral extent for a spatial ( pixel ) location at a time , i.e. , @xmath0 $ ] .
then , the compressed measurements @xmath1 are sent to a fusion station that will recover the original 3d datacube by utilizing a cs reconstruction algorithm . by taking the highly correlated hyperspectral data vectors @xmath2 to admit a sparse representation in some orthonormal basis , e.g. , dct basis or wavelet basis , we have , @xmath3 where sparse signal vectors @xmath4 .
@xmath5 is a certain orthonormal basis matrix .
then we obtain the typical cs formulation as follows @xmath6where @xmath7\in\mathbb{r}^{m_t\times n}$ ] , @xmath8 is the measurement output vector , and @xmath9 is an additive noise vector with unknown variance @xmath10 .
[ section2.3 ] as mentioned in the introduction , while the original hyperspectral data can be reconstructed by using conventional cs recovery algorithms , it is possible to achieve a much better recovery performance by applying the l&s model to further exploit the structural dependencies between the values and locations of the coefficients of the sparse signal vectors @xmath11 .
the main reason that we consider @xmath12 as a l&s matrix is two - fold .
first , images from different spectral bands enjoy similar natural image statistics , and hence can be joint - sparse in a wavelet / dct basis ; second , a limited number of unique materials in a scenes implies that spectral signatures across pixels can be stacked to form a matrix that is often low - rank @xcite . to precisely achieve the benefits of the l&s model and reconstruct the original hyperspectral data from a bayesian point of view , here we propose an accurate probabilistic model by performing a structured l&s factorization for @xmath12 as
@xmath13where the diagonal matrix @xmath14 is the sparsity pattern matrix of the signals with the support indicates @xmath15 .
we refer to @xmath16 as the sparsity level of @xmath12 .
@xmath17\in\mathbb{r}^{n\times r}$ ] and @xmath18\in\mathbb{r}^{r\times t}$ ] are obtained from the low - rank matrix factorization of @xmath19 , which is the amplitude matrix of @xmath12 . for a joint - sparse matrix @xmath20 and an arbitrary matrix @xmath21
, this factorization implies that @xmath12 is a simultaneous low - rank ( @xmath22 ) and joint - sparse matrix with rank @xmath23 , where all sparse signal vectors @xmath24 share a common support with sparsity level @xmath25 .
assuming independent entries for @xmath26 , @xmath27 , and @xmath21 , the separable probability density functions ( pdfs ) of @xmath28 and @xmath21 become @xmath29 @xmath30 where both @xmath31 and @xmath32 are assumed to be i.i.d .
gaussian with unknown mean @xmath33 and variance @xmath34 .
in particular , we assume @xmath32 follow i.i.d .
gaussian distribution with zero mean and unit variance , i.e. , @xmath35 , to avoid ambiguity and the unnecessary model parameters update . as @xmath36
are treated as i.i.d .
bernoulli random variables with @xmath37 , the sparse coefficients , @xmath38 , become i.i.d .
bernoulli - gaussian ( bg ) , i.e. , the marginal pdf @xmath39 where @xmath40 is the dirac delta function .
furthermore , due to the assumption of adaptive gaussian noise in ( [ cs ] ) , the likelihood function of @xmath41 is known and separable , i.e. , @xmath42where the measurement @xmath43 . , @xmath44 , @xmath45 , and @xmath46.,title="fig:",width=264 ] +
[ section3 ] with the problem formulation in ( [ cs ] ) and ( [ signalmodel ] ) , our proposed method is to maximize the posterior joint distribution , i.e. , @xmath47 @xmath48 @xmath49 @xmath50 where @xmath51 denotes equality up to a normalizing constant scale factor .
this posterior distribution can be represented with a factor graph shown in fig .
[ fig_l&samp ] , where circles denote random variables and squares denote posterior factors based on belief propagation @xcite .
each factor node represents the conditional probability distribution between all variable nodes it connected .
@xmath52 vertical planes ( parallel to @xmath53 and @xmath54 axes ) exploit the linear measurement structure @xmath55 ( detailed in fig .
[ fig_mgamp ] ) , while the remaining part of fig .
[ fig_l&samp ] further exploits the l&s structure @xmath56 . to bypass the intractable inference problem of marginalizing ( [ pjd ] )
, we propose to solve an alternative problem that consists of two sub - problems that mainly require local information to complete their tasks @xcite .
correspondingly , our proposed algorithm is divided into two phases : i ) the _ multiple generalized approximate message passing _
( m - gamp ) phase ; ii ) the _ low - rankness and sparsity pattern decoding _ ( l&spd ) phase . owing to this ,
an efficient turbo amp " iterative framework @xcite is used , that iteratively updates one of the phases beliefs , and passes the beliefs to another phase , and vice versa , repeating until both phases converge . , @xmath44 , @xmath45 , and @xmath46.,title="fig:",width=238 ] + [ subsection3_1 ] in each frame @xmath57 of the m - gamp phase
, we apply the _ generalized approximate message passing _ ( gamp ) approach @xcite in parallel for the linear inference problem : estimate the vector @xmath58 from the observation @xmath59 , as shown in fig .
[ fig_mgamp ] . specifically , the gamp computes the approximated posteriors on @xmath60 as @xcite , @xcite @xmath61 @xmath62where @xmath63 denotes a message passed from node @xmath64 to the adjacent node @xmath65 in the factor graph .
the parameters @xmath66 , and @xmath67 are obtained after the gamp iteration converges .
for the prior distribution of @xmath68 , i.e. , @xmath69 used in ( [ xpost ] ) , we can assume bg prior pdf @xmath70 where @xmath71 and @xmath72 are the active - coefficient mean and the active - coefficient variance , respectively , of the variable @xmath68 .
it is worth mentioning that the prior parameters @xmath73 and @xmath74 , are only initialized to agnostic values at the beginning of the l&s - amp algorithm ( e.g. , @xmath75 ) , then iteratively updated according to the message passed from the l&spd phase .
this process will be detailed in next subsection .
then the minimum - mean - squared error ( mmse ) estimation of @xmath76 is facilitated by the following prior - dependent integrals @xmath77 [ subsection3_2 ] in the l&spd phase , to exploit the l&s structure , we employ the recently proposed _
bilinear generalized approximate message passing _
( big - amp ) approach @xcite to a variant of the pca problem : estimate the matrices @xmath28 and @xmath12 from an observation @xmath78\in\mathbb{r}^{n\times t}$ ] which is the posterior estimation of their product @xmath79 obtained form the m - gamp phase in ( [ xmarginal ] ) . in particular , the big - amp @xcite obtains the approximately gaussian posterior messages @xmath80 as @xmath81\ ] ] @xmath82 where the parameters @xmath83 and @xmath84 are obtained after the big - amp iteration converges .
the prior distribution of @xmath85 , i.e. , @xmath86 used in ( [ xpost1 ] ) , comes from the posterior message of @xmath87 given the observation @xmath59 in the m - gamp phase .
the prior distribution of @xmath69 used in ( [ xpost ] ) comes from the posterior message of @xmath68 given the matrix factorization @xmath88 in the l&spd phase . to enable effective implementation of turbo amp " iteration , given the construction of the factor graph in fig .
[ fig_l&samp ] , the sum - product algorithm ( spa ) @xcite implies that , @xmath89 @xmath90 @xmath91 . \vspace{-4pt}\]]comparing ( [ xprior_spa ] ) and ( [ xprior_spa1 ] ) with ( [ xpost ] ) and ( [ xpost1 ] ) , respectively , we have @xmath92 @xmath93 thus , the parameters @xmath94 computed during the final iteration of the m - gamp phase , are treated as the prior parameters of @xmath12 in the l&spd phase .
conversely , the parameters @xmath95 and @xmath96 computed during the final iteration of the l&spd phase are in turn used as the prior parameters of @xmath12 in the m - gamp phase in ( [ xmarginal ] ) .
in addition , to further exploiting the joint - sparsity of @xmath97 , we use the local support estimate @xmath98 instead of the common sparsity rate @xmath99 in ( [ xmarginal ] ) .
then , by applying the spa in the m - gamp phase , we get @xmath100where the posterior local support probability @xmath101 [ subsection3_3 ] beginning at the initial inter - phase iteration index , @xmath102 , the l&s - amp algorithm first performs the m - gamp phase with the initial prior parameters @xmath103 in ( [ xmarginal ] )
. then the converged outgoing messages @xmath104 are treated as prior parameters in the l&spd phase . then the converged messages @xmath73 and @xmath105 obtained from the l&spd phase , along with the updated beliefs @xmath106 in ( [ lambdain ] ) , are used for the m - gamp phase at inter - phase iteration @xmath107 .
this procedure continues until either a stopping condition or a maximum number of allowable iterations is reached .
then we obtain the posterior mean estimates @xmath108 computed in ( [ xestimate ] ) .
furthermore , we tune our prior and likelihood parameters @xmath109 using expectation - maximization @xcite,@xcite , and estimate the rank @xmath110 using a rank selection strategy based on the penalized log - likelihood maximization in @xcite .
in addition , we recommend initializing l&s - amp using @xmath111 , @xmath112 and @xmath113 .
[ section4 ] in this section , we present real data results to compare the performance of the proposed l&s - amp algorithm with prior state - of - art ppxa , ra - ormp @xcite , sa - music @xcite , and t - msbl @xcite algorithms .
we evaluate the performance of the algorithms on two real hyperspectral datasets : 1 ) an urban dataset acquired over the university of houston , with 144 spectral bands , @xmath114 pixels , and a spatial resolution of @xmath115 .
2 ) an agricultural dataset acquired over the salinas valley in california .
the dataset has a spatial resolution of @xmath116 and consists of @xmath117 spectral bands with each band corresponding to an image with @xmath118 pixels .
we assume that the l&s signal matrix @xmath119 is obtained using pixel - based acquisition , so that @xmath52 denotes the number of pixels and @xmath120 denotes the number of spectral band .
the dct matrix is used as the sparsifying matrix @xmath121 , and gaussian noise is added to achieve snr @xmath122 db .
it is worth noting that , for the sake of comparion , different random gaussian measurement matrices @xmath123 are used .
also note that t - msbl , ra - ormp , and sa - music are derived only for the common measurement matrix case . for the recovery of the urban dataset ( left ) and agriculture - oriented dataset ( right).,title="fig:",width=302 ] + fig .
[ cnmse ] plots the column - averaged normalized mse ( cnmse ) versus the compressive ratio @xmath124 on the two real datasets .
the cnmse is defined as @xmath125 , where @xmath126 is an estimate of @xmath12 . from the figure , we observe that the proposed algorithm outperforms all the other algorithms in terms of cnmse , e.g. , in fig . [ cnmse].(b )
, we note that l&s - amp achieves nearly 3db reconstruction gain than the other algorithms at @xmath127 .
in addition , a plus - minus sign ( @xmath128 ) is used ( i.e. , l&s - amp@xmath129 ) to denote the case of using random @xmath130 measurement matrices , which are easy to implement in dmd , and can significantly reduce the burden of storage . some visual results of the recovered hyperspectral images by using different algorithms are presented in fig .
[ visual ] . as expected ,
our proposed algorithm preserves more fine details and much sharper edges , and shows much clearer and better visual results than the other competing methods .
is fixed to 0.243 , and other simulation parameters remain unchanged . the whole scene is partitioned into a sequence of sub - scenes to enable parallel processing . ,
title="fig:",width=317 ] +
[ section5 ] in this paper , we studied joint cs reconstruction of spatially and spectrally correlated hyperspectral data acquired , assuming that the hyperspectral signal matrix satisfies the joint - sparse model with a lower rank than the sparsity level , i.e. , the l&s model .
we proposed an amp - based algorithm for recovering the signal matrix with the l&s model while exploiting the structured sparsity and the amplitude correlation of the data .
the numerical results were presented to confirm the performance advantage of our algorithm . 1 a. plaza , j. a. benediktsson , j. boardman , j. brazile , l. bruzzone , g. camps - valls , j. chanussot , m. fauvel , p. gamba , j. gualtieri , m. marconcini , j. c. tilton , and g. trianni , recent advances in techniques for hyperspectral image processing , " _ remote sens . environment , _ vol . 113 , no . 1 , pp .
110122 , sep . 2009 .
j. m. bioucas - dias , a. plaza , g. camps - valls , p. scheunders , n. m. nasrabadi , and j. chanussot , hyperspectral remote sensing data analysis and future challenges , " _ ieee geoscience and remote sens . mag .
, _ vol . 1 , no
. 2 , pp . 636 , jun . 2013 .
e. j. cands , j. romberg , and t. tao , robust uncertainty principles : exact signal reconstruction from highly incomplete frequency information , " _ ieee trans .
theory , _ vol .
489509 , feb .
m. f. duarte , m. a. davenport , d. takhar , j. n. laska , t. sun , k. e. kelly , and r. g. baraniuk , single - pixel imaging via compressive sampling , " _ ieee signal process .
2 , pp : 8391 , mar .
2008 .
m. golbabaee and p. vandergheynst , hyperspectral image compressed sensing via low - rank and joint - sparse matrix recovery , " in _ proc .
conf . on acoustics , speech , and
signal proces .
( icassp ) , _ pp : 27412744 , kyoto , japan , mar .
2012 .
r. m. willett , m. f. duarte , m. a. davenport , and r. g. baraniuk , sparsity and structure in hyperspectral imaging : sensing , reconstruction , and target detection , " _ ieee signal process . mag .
1 , pp : 116126 , jan . 2014 . z. zhang and b. d. rao , sparse signal recovery with temporally correlated source vectors using sparse bayesian learning , " _ ieee j. select .
topics signal process .
, _ vol . 5 , no .
5 , pp . 912926 , sep . | this paper considers a compressive sensing ( cs ) approach for hyperspectral data acquisition , which results in a practical compression ratio substantially higher than the state - of - the - art . applying simultaneous low - rank and joint - sparse ( l&s ) model to the hyperspectral data , we propose a novel algorithm to joint reconstruction of hyperspectral data based on loopy belief propagation that enables the exploitation of both structured sparsity and amplitude correlations in the data .
experimental results with real hyperspectral datasets demonstrate that the proposed algorithm outperforms the state - of - the - art cs - based solutions with substantial reductions in reconstruction error . compressive hyperspectral imaging , low - rank and joint - sparse , compressive sensing , approximate message passing |
recent advances in chemical synthesis have unveiled more and more the importance of soft - matter systems , such as dispersions of colloidal particles , polymers , and their combinations . besides their practical interest
, these new developments have opened up new theoretical avenues in ( at least ) two instances .
firstly , it is possible to experimentally fine - tune the details of interactions ( range , strength , ) , making these systems a unique laboratory for testing highly simplified models within an effective - interaction approach where the microscopic degrees of freedom are integrated out in favor of renormalized macroparticle interactions .
secondly , they offer the possibility of exploring new types of equilibrium phase behaviors not present in the simple - fluid paradigm .
as early as in 1989 , marquest and witten@xcite suggested that the experimentally observed crystallization in some copolymer micellar systems could be rationalized on the basis of a bounded interaction , that is , an interaction that does not diverge at the origin .
successive theoretical work showed that this class of bounded or ultra - soft potentials naturally arises as effective interactions between the centers of mass of many soft and flexible macromolecules , such as polymer chains , dendrimers , star polymers , etc .
( see , e.g. , ref . for a reference on the subject ) .
two well - studied cases belonging to the above class are the gaussian core model ( gcm ) introduced by stillinger@xcite and the penetrable - sphere ( ps ) model introduced in refs . , whose freezing transition turns out to display rather exotic features with no analogue in the atomistic fluid realm . in the present paper
, we shall consider a close relative of the ps model , first introduced in ref .
, denoted as the penetrable - square - well ( psw ) model , where a short - range attractive tail is added to the ps model just outside the core region . in the limit of infinite repulsive energy ,
the ps and psw models reduce to the usual hard - sphere ( hs ) and square - well ( sw ) models , respectively .
an additional interesting feature common to both ps and psw , as well as to all bounded potentials , is the fact that , even when confined to one - dimensional systems , they may exhibit a non - trivial phase diagram due to the penetrability which prevents an exact analytical solution .
this is because particles can not be lined up on a line with a well - defined and fixed ordering in view of the possibility of reciprocal interpenetration ( with some positive energy cost ) , thus lacking an essential ingredient allowing for an exact solution in the respective hard - core counterparts ( hs and sw ) .
it is then particularly useful to discuss some motivated approximations in the one - dimensional model which can then be benchmarked against numerical simulations and subsequently exploited in the much more complex three - dimensional case .
the aim of the present paper is to complete a study on the one - dimensional psw model started in refs . , as well as the general results presented in ref . which are particularly relevant in the present context
in the first paper of the series,@xcite we introduced the model and discussed the range of stability in terms of the attractive versus repulsive energy scale .
we provided , in addition , exact analytical results in the low - density limit ( second order in the radial distribution function and fourth order in the virial expansion ) and a detailed study of the percus yevick and hypernetted - chain integral equations .
these were used in the following paper@xcite to propose a low - penetrability approximation at finite density which was then tested against numerical simulation .
this low - penetrability approximation is expected to break down in the opposite regime , namely when temperatures and densities are such that particles easily interpenetrate each other . in the present paper , we address this latter regime by proposing a complementary approximation ( the high - penetrability approximation ) and discussing its range of validity and the relationship with the low - penetrability regime .
note that a similar matching of the low- and high - penetrability approximation has already been carried out by two of the present authors in the framework of the ps model.@xcite it turns out that the high - penetrability approximation in the context of bounded potentials coincides with the linearized debye
hckel classical approximation originally introduced in the framework of the coulomb potential.@xcite it has been recently shown@xcite that two of the three standard routes to thermodynamics ( the energy and the virial routes ) are automatically consistent within the linearized debye hckel approximation , for any potential and dimensionality .
this means that a deviation from the third standard route to thermodynamics ( the compressibility route ) can be exploited to assess the degree of reliability of the high - penetrability ( or linearized debye
hckel ) approximation .
this is indeed discussed in the present paper , where we also discuss the full hierarchy of approximations ranging from the full debye
hckel approximation to the simplest mean - spherical approximation . in view of the boundness of the potential
, the usual van hove no - go theorem@xcite on the absence of phase transitions in certain one - dimensional fluids does not hold .
it is then natural to ask whether a phase transition occurs in the one - dimensional psw fluid by noting that the addition of an attractive tail to the pair potential of the ps model extends the question to the fluid - fluid transition , in addition to the fluid - solid transition possible even within the ps model . in the present paper
we confine our attention to the fluid - fluid case only and discuss this possibility using both analytical arguments and state - of - the - art numerical simulations.@xcite our results are compatible with the absence of such a transition , as we shall discuss .
this is also supported by recent analytical results@xcite using a methodology devised for one - dimensional models with long - range interactions.@xcite we discuss possible reasons for this and a plausible scenario for the three - dimensional case .
finally , we note that the approach to a critical point is frequently anticipated by the so - called fisher
widom line@xcite marking the borderline between a region with oscillatory behavior in the long - range domain of the correlation function ( above the fisher widom line ) and a region of exponential decay .
we discuss the location of this line within the high - penetrability approximation and again the matching of this result with that stemming from the low - penetrability approximation .
the structure of the paper is as follows .
we define the psw model in sec .
[ sec : model ] .
we then construct the high - penetrability approximation in sec .
[ sec : hpa ] and in sec .
[ sec : rhpa ] we discuss some approximations related to it .
section [ sec : eos ] contains a discussion on the routes to thermodynamics , as predicted by the high - penetrability approximation .
the structure predicted by the approximation is compared with monte carlo data in sec . [
sec : structure ] .
the fisher widom line and the possibility of a fluid - fluid transition are discussed in sec . [
sec : fw ] .
the paper ends with some concluding remarks in sec .
[ sec : conclusions ] .
the penetrable - square - well ( psw ) model is defined by the following pair potential@xcite @xmath0 where @xmath1 is the well width and @xmath2 and @xmath3 are two positive constants accounting for the repulsive and attractive parts of the potential , respectively . in the following , we shall restrict our analysis to the case @xmath4 and @xmath5 , where we know the one - dimensional model to be stable with a well defined thermodynamic limit.@xcite it is shown in appendix [ app : ruelle ] that , more generally , the one - dimensional psw model is guaranteed to be stable if @xmath6 , where @xmath7 is the integer part of @xmath8 . for lower values of @xmath2 the model may or may not be stable and we will come back to this point in section [ sec : fw ] .
an important role in the following is played by the corresponding mayer function @xmath9,\ ] ] where @xmath10 is the parameter measuring the degree of penetrability varying between 0 ( free penetrability ) and 1 ( impenetrability ) , while @xmath11 measures the strength of the well depth . here
@xmath12 with @xmath13 the temperature and @xmath14 the boltzmann constant , @xmath15 is the mayer function for hard spheres , and @xmath16 is the heaviside step function
. a detailed discussion of the limiting cases of the psw model can be found in ref . .
here we merely note that the psw mayer function @xmath17 is immediately related to the usual sw mayer function by @xmath18 @xmath19~,\ ] ] where we have introduced the ratio @xmath20 . at a given value of @xmath21 , @xmath22 increases quasi - linearly with @xmath23 , its minimum value @xmath24 corresponding to @xmath25 .
in ref . we discussed a low - penetrability approximation ( lpa ) to the psw model . within this approximation , one assumes @xmath26 so that the repulsive barrier @xmath2 is sufficiently higher than the thermal energy @xmath27 , penetrability is small , and the system is almost a hard - core one .
the advantage of this theoretical scheme is that one can use the general recipe leading to the exact solution for the one - dimensional sw problem in fact , valid for any potential with a hard core and short - range attractions and perform some ad hoc adjustments to ensure that some basic physical conditions on the radial distribution function @xmath28 are satisfied .
comparison with monte carlo ( mc ) simulations showed a good behavior of the lpa even for @xmath29 ( @xmath30 ) , provided the density was moderate ( @xmath31 ) .
the opposite limit @xmath32 is also inherently interesting for several reasons . from a physical viewpoint
this amounts to starting from the ideal gas limit @xmath33 ( one of the common reference systems for simple fluids ) and progressively building up interactions by increasing @xmath34 .
an additional mathematical advantage stems from the simple observation@xcite that in the ( exact ) cluster expansion of @xmath28 only the dominant chain diagrams need to be retained at all orders , thus leading to the possibility of an exact summation of those leading contributions . as we shall see
shortly , this is in fact a procedure known as the debye
hckel approximation in the context of charged fluids.@xcite our main goal is the computation of the cavity function @xmath35 , from which one can immediately compute the radial distribution function ( rdf ) @xmath36 $ ] . in the psw case one
then has from eq .
( [ model : eq3 ] ) @xmath37 where @xmath38 and , in conformity with previous work,@xcite we have redefined all lengths in units of @xmath39 so we set @xmath40 in most of the following equations . as shown in ref . for the ps case , the exact form of the psw cavity function in the limit
@xmath33 at finite @xmath20 and @xmath41 is @xmath42 where the function @xmath43 is defined through its fourier transform @xmath44 @xmath45 being the fourier transform of @xmath46 .
note that in the limit @xmath33 one has @xmath47 and @xmath48 .
generalizing an analogous approximation in the context of the ps model,@xcite our high - penetrability approximation ( hpa ) consists of assuming eqs .
( [ hpa : eq3 ] ) and ( [ hpa : eq4 ] ) for finite , but small , values of @xmath34 .
it is worth noting that the combination of the two expressions ( [ hpa : eq3 ] ) and ( [ hpa : eq4 ] ) defines what is usually referred to , in a different context , as the linearized debye
hckel ( ldh ) approximation,@xcite and this will be further elaborated below .
equations ( [ hpa : eq3 ] ) and ( [ hpa : eq4 ] ) hold for any dimensionality . in the specific one - dimensional case , and taking into account eq .
, we have @xmath49~.\end{aligned}\ ] ] the function @xmath43 can be numerically evaluated in real space by fourier inversion as @xmath50 an explicit expression for the density expansion of @xmath43 within the hpa is reported in eqs .
( [ appa : eq6 ] ) and ( [ appa : eq7 ] ) of appendix [ app : appa ] , where the radius of convergence of the expansion is also analyzed . from eqs .
( [ model : eq3 ] ) , ( [ hpa : eq2 ] ) , and ( [ hpa : eq3 ] ) the total correlation function , @xmath51 , within the hpa is easily obtained as @xmath52+\gamma_rf_{{\text{sw}}}\left(r\right)~,\end{aligned}\ ] ] or in fourier space , @xmath53 from this equation it is straightforward to get the structure factor @xmath54 and the fourier transform of the direct correlation function @xmath55 . the zero wavenumber value of the structure factor is @xmath56 where we have taken into account that @xmath57 .
this completes the calculation of the correlation functions within the hpa .
as anticipated in section [ sec : hpa ] , the hpa is the exact equivalent to the well - known ldh approximation , which is widely used in the context of charged simple fluids.@xcite the latter is actually an intermediate step of a hierarchy of successive approximations ranging from the simplest mean - spherical approximation ( msa ) to the full non - linear version of the debye
hckel approximation ( see ref . [ 11 ] in ref . for a discussion on this point ) . for the psw model
and more generally for any bounded potential the ldh approximation ( the hpa in the present language ) is particularly relevant in view of the fact that one can make @xmath17 arbitrarily small by letting @xmath58 , thus justifying the approximation of neglecting non - chain diagrams .
it is then interesting to check the performance of the other approximations included in the aforementioned class , which will be translated in the present context for simplicity . on top of the hierarchy of approximations
there is the non - linear hpa ( nlhpa ) @xmath59 which is equivalent to the non - linear debye
hckel approximation , as remarked .
the hpa , eq .
, is obtained upon linearizing the exponential , an approximation valid again in the limit @xmath60 .
an additional approximation denoted here as the modified hpa ( mhpa) can be considered with the help of eq .
( [ hpa : eq5 ] ) by neglecting the quadratic term in @xmath34 .
this yields @xmath61~.\end{aligned}\ ] ] this is equivalent to keeping only the first term on the right - hand side of eq . , which implies @xmath62 or , in real space , @xmath63 the lowest rank in the hierarchy is occupied by the msa , which is obtained from eq .
( [ rhpa : eq3 ] ) upon linearization of the mayer function @xmath17 , @xmath64 since @xmath43 is a convolution , it must be continuous at @xmath65 and @xmath66 .
it follows that the approximations with a continuous cavity function at @xmath65 and @xmath66 are nlhpa and hpa .
for instance , in the mhpa the cavity function is @xmath67 $ ] , so that one has @xmath68 and @xmath69 .
it has been shown in ref . that the virial and energy routes to thermodynamics ( to be discussed below ) are consistent one another within the hpa , for any potential and any dimensionality .
a similar statement holds true for soft potentials within the msa.@xcite this clearly includes the psw potential in both cases .
it is interesting to make contact with previous work carried out by likos et al .
@xcite on a general class of unbounded potentials which are free of attractive parts , thus resulting particularly useful in the context of the fluid - solid transition . in ref . , the msa given in eq .
( [ rhpa : eq4 ] ) along with the spinodal instability to be discussed in detail in section [ sec : fw ] , have been introduced for a general class of unbounded potentials including the ps as a particular case .
this has been further elaborated and extended to include the gcm in ref . .
in both cases , the authors discuss directly the three - dimensional case , so that a direct comparison with the present work can not be drawn at the present stage , but they also provide a detailed discussion of various approximations , within the general framework of density functional theory , that provide a unified framework where even the present model could be included .
given an approximate solution of a fluid model there are several routes to the equation of state which , in general , give different results .
the most common are three:@xcite the virial , the compressibility , and the energy route .
the consistency of the outcome of these different routes can be regarded as an assessment on the soundness of the approximation .
for some particular approximations it may also happen that the consistency of two of the three routes are automatically enforced ( see ref . and references therein for a detailed discussion on this point ) .
this is the case of the hpa , where the virial and energy routes coincide , as anticipated .
hence , the consistency with the compressibility route will provide a rough estimate of the regime of validity of the hpa within a phase diagram for the psw potential .
let us briefly recall@xcite the methods to compute the compressibility factor , @xmath70 , associated with the three different routes .
the virial route is defined by @xmath71 which , using standard manipulations,@xcite yields @xmath72.\end{aligned}\ ] ] thus the problem is reduced to the computation of the cavity function @xmath73 which , in the present context , follows from eqs .
( [ hpa : eq3 ] ) and ( [ w(r ) ] ) . next
, we consider the compressibility route , @xmath74 where the integral can be readily evaluated with the help of eq .
( [ s(0 ) ] ) . regarding the energy route , we start from the internal energy per particle @xmath75 in the above equation , the expressions given by eqs .
( [ model : eq1 ] ) and ( [ hpa : eq2 ] ) have been used . in order to obtain @xmath76 from @xmath77
we exploit the following standard thermodynamic identity@xcite @xmath78 thus leading to @xmath79 we have used the exact consistency between the virial and energy routes within the hpa as a test of the numerical calculations . as computed from the virial(-energy ) route to the pressure for @xmath80 and @xmath81 .
results for three different reduced temperatures @xmath82 are displayed .
[ fig : fig1 ] ] figure [ fig : fig1 ] depicts the results of the virial and hence energy , as remarked route under the condition @xmath80 , which constitutes the borderline range of stability of the one - dimensional psw model with @xmath4.@xcite under this demanding condition , we have considered three reduced temperatures from @xmath83 to @xmath84 , whereas the width of the well has been fixed to the value @xmath81 .
we remark that @xmath85 is not limited in values from above due to the boundness of the potential .
the clear downturn of all three curves for sufficiently large reduced density @xmath85 is a consequence of the existence of a maximum density @xmath86 [ see eq . ] , beyond which the hpa breaks down , as described at the end of appendix [ app : appa ] . in particular , the values of the maximum density for @xmath80 and @xmath81 are @xmath87 , @xmath88 , and @xmath89 at @xmath90 , @xmath91 , and @xmath92 , respectively , in agreement with fig .
[ fig : fig1 ] . from the virial(-energy ) and the compressibility route to the pressure .
the circles represent mc simulation results .
chosen parameters are @xmath80 , @xmath93 , and @xmath94 .
[ fig : fig2 ] ] we compare in fig . [ fig : fig2 ] the results from the virial(-energy ) and the compressibility routes with mc simulations@xcite for an intermediate value of the reduced temperature ( @xmath95 ) and other parameters as before .
rather interestingly , the virial(-energy ) route appears to reproduce rather well the numerical simulation results up to the region where the artificial downward behavior shows up , whereas the compressibility route begins to deviate for densities @xmath96 .
versus reduced temperature @xmath97 plane . here
the curves ( a)(d ) correspond to @xmath98 , @xmath99 , @xmath100 , and @xmath101 , respectively .
the region below each curve represents states where the relative deviation between the virial(-energy ) route and the compressibility one is smaller than @xmath102 and hence regarded as reliable .
the inset shows the curves in the @xmath85 versus @xmath103 plane for @xmath104 ( dashed line ) and @xmath99 ( solid line ) .
[ fig : fig3 ] ] as the temperature increases the hpa theory clearly remains a good approximation for a larger range of densities .
we can naturally measure this by the requirement that virial(-energy ) and compressibility routes are consistent within a few percent .
this is indeed shown in fig .
[ fig : fig3 ] , where we depict a transition line separating a `` reliable '' from an `` unreliable '' regime , as measured by the relative deviation of the two routes ( here taken to be @xmath102 ) , for four choices of the model parameters : @xmath104 , @xmath101 , @xmath99 , and @xmath105 .
the value @xmath106 is frequently used in the sw counterpart.@xcite we observe that the region @xmath107 where the hpa is reliable is hardly dependent on @xmath8 [ compare curves ( a)&(b ) and ( c)&(d ) in fig . [ fig : fig3 ] ] . on the other hand , at given values of @xmath8 and @xmath108 , the range @xmath107 decreases with increasing @xmath21 [ compare curves ( a)&(d ) and ( b)&(c ) in fig .
[ fig : fig3 ] ] , as expected . however , this effect is much less important if the increase of @xmath21 takes place at fixed @xmath109 ( see inset of fig .
[ fig : fig3 ] ) .
it is interesting to note that , as illustrated by fig .
[ fig : fig2 ] , the hpa virial route keeps being reliable up to a certain density higher than @xmath110 . versus reduced temperature @xmath97 plane .
here @xmath21=5 and @xmath81 .
the labels lpa , hpa , and lpa+hpa indicate the regions where only the lpa , only the hpa , or both approximations are reliable , respectively .
[ fig : fig3b ] ] as said above , in ref .
we introduced a lpa that was accurate for states where the penetrability effects were low or moderate .
the lpa is complemented by the hpa presented in this paper .
it is then interesting to compare the regions where each approximation can be considered reliable according to the same criterion as in fig .
[ fig : fig3 ] .
this is shown in fig .
[ fig : fig3b ] for the case @xmath111 and @xmath81 .
the two transition lines split the plane into four regions : a region where only the lpa is reliable , a region where only the hpa is reliable , a region where both approximations are reliable ( and provide equivalent results ) , and a region where none of them is sufficiently good .
the latter region shrinks as @xmath21 decreases ( thanks to the hpa ) or @xmath8 decreases ( thanks to the lpa ) .
as an additional test of the soundness of the hpa , we also study the rdf @xmath112 , which can easily be obtained from eqs . and , or equivalently from eq .
( [ hpa : eq5 ] ) , once the auxiliary function @xmath43 has been determined . for a sufficiently high temperature ( and hence high penetrability ) , the hpa is clearly well performing , as can be inferred from fig .
[ fig : fig4 ] , when compared with standard nvt mc results . here
we have considered the same parameters as in the preceding section ( @xmath80 and @xmath81 ) at a corresponding high - temperature value @xmath113 and a density @xmath114 where overlappings are unavoidable . under these conditions
, there is no visible difference among the various approximations considered in section [ sec : rhpa ] .
the excellent performance of the hpa observed in fig .
[ fig : fig4 ] agrees with the reliability criterion of fig . [
fig : fig3 ] since the state @xmath113 and @xmath114 is well below the curve ( c ) corresponding to @xmath80 and @xmath81 . as a function of @xmath115 with @xmath116 , @xmath81 , @xmath113 , and @xmath117 .
predictions from the hpa ( solid line ) are compared with mc results ( circles ) .
[ fig : fig4 ] ] vs @xmath115 .
chosen parameters are @xmath116 , @xmath81 , and @xmath117 .
mc results ( circles ) are compared with the hpa ( solid line ) , the nlhpa ( dashed line ) , the mhpa ( short dashed line ) , and the msa ( long dashed line ) .
the two different panels refer to different reduced temperatures : @xmath118 ( top panel ) and @xmath119 ( bottom panel ) .
note that both panels are drawn within the same scale .
[ fig : fig5 ] ] as we cool down , significant differences among various approximations ( hpa , nlhpa , mhpa , and msa ) begin to appear , as depicted in fig .
[ fig : fig5 ] , where results corresponding to temperatures @xmath120 ( top panel ) and @xmath121 ( bottom panel ) are reported within the same scale .
the state @xmath120 and @xmath114 is still lying in the reliable region of fig .
[ fig : fig3 ] , but close to the boundary line ( c ) , while the state @xmath121 and @xmath114 is clearly outside that region . in the case
@xmath120 the hpa and its three variants are practically indistinguishable , except in the region @xmath122 , which is very important to describe the correct thermodynamic behavior , where the best agreement with mc data corresponds to the nlhpa , followed by the hpa .
the two approximations that do not preserve the continuity of the cavity functions ( mhpa and msa ) overestimate the jump at @xmath123 . in the lower temperature case @xmath121 all the approximations overestimate the oscillations of the rdf .
interestingly , the hpa captures quite well the values of @xmath28 near the origin .
the worst overall behavior corresponds again to the msa , which even predicts negative values of @xmath28 for @xmath124 .
vs @xmath115 .
here fixed parameters are @xmath125 and @xmath106 .
mc results ( circles ) are compared with the hpa ( solid line ) , the nlhpa ( dashed line ) , the mhpa ( short dashed line ) , and the msa ( long dashed line ) .
the top panel refers to the state @xmath118 and @xmath117 , whereas the bottom panel refers to @xmath126 and @xmath127 . in the bottom panel
the msa is not depicted .
again both panels are on the same scale .
[ fig : fig6 ] ] vs @xmath115 .
chosen parameters are @xmath111 , @xmath81 , @xmath128 , and @xmath129 .
mc results ( circles ) are compared with the hpa ( solid line ) and the lpa ( dashed line ) .
[ fig : fig6b ] ] additional insights can be obtained by decreasing the range of interactions , in close analogy with what we considered in previous work for the complementary lpa.@xcite this is reported in fig .
[ fig : fig6 ] for the case @xmath111 and @xmath106 with @xmath130 ( top panel ) and @xmath131 ( bottom panel ) .
again we stress that the same scale is used for both panels in order to emphasize the effect of lowering the temperature .
clearly this is a more demanding situation .
in fact , both states are above curve ( a ) of fig .
[ fig : fig3 ] and thus outside the corresponding reliability region .
therefore , clear deviations from mc results appear in all considered approximations , especially in the lower temperature case @xmath128 ( bottom panel ) . yet
, the hpa is still a reasonably good approximation that follows the main qualitative features of the correct @xmath28 . in the higher temperature case @xmath120
the only noticeable limitations of the hpa practically take place near the origin , this deficiency being largely corrected by the mhpa . to conclude this section , it is worthwhile comparing the two complementary approaches hpa and lpa at a case where both are expected to be reliable , according to the diagram of fig .
[ fig : fig3b ] .
this is done in fig .
[ fig : fig6b ] for @xmath111 , @xmath81 , @xmath128 , and @xmath129 .
we observe that both approximations agree well each other and with mc data , except in the region @xmath132 , where the hpa rdf presents an artificial curvature .
we now turn to an interesting point raised in previous work,@xcite namely the question of whether the model can display a phase transition in spite of its one - dimensional character .
the existence of general theorems all essentially based on the original van hove s result@xcite on the absence of phase transitions for a large class of one - dimensional models with short - range interactions is well established.@xcite psw and ps models , however , do not belong to the class for which these general theorems hold .
this is because boundness allows multiple partial ( or even total ) overlapping at some energy cost , thus rendering invalid the arguments used in the aforementioned theorems .
on the other hand , none of these theorems provide a general guideline to understand whether a one - dimensional model may or may not display a non - trivial phase transition , and one has then to rely on the specificity of each model .
as discussed in our previous work,@xcite it is instructive to first address the simpler question of the location of the fisher widom ( fw ) line .
this is a line separating two different regimes for the large - distance behavior of the rdf @xmath28 in the presence of competing repulsive / attractive interactions.@xcite the rationale behind the fw line is that on approaching the critical points where attractions become more and more effective , the behavior of correlation functions must switch from oscillatory ( characteristic of repulsive interactions ) to exponential with a well defined correlation length @xmath133 . in the previous work
, we analyzed the location of this line for psw within the lpa .
here we extend this analysis to the hpa regime and discuss the compatibility of the two results .
let us first briefly recall the main points of the analysis , referring to ref . for details . from eq .
( [ hpa : eq2 ] ) we note that the asymptotic behavior of @xmath28 is the same as that of @xmath73 . in view of eq .
( [ hpa : eq3 ] ) , this is hence related to @xmath43 , whose asymptotic behavior is governed by the pair of conjugate poles of @xmath134 with an imaginary part closest to the origin . if the real part of the pair is zero , the decay is monotonic and oscillatory otherwise .
according to eq . , the poles of @xmath134 are given by @xmath135 let @xmath136 be the imaginary pole and @xmath137 be the pole with the imaginary part closest to the origin .
the fw line is determined by the condition @xmath138 .
this gives , at a given temperature , three equations in the three unknowns @xmath139 , @xmath140 , and @xmath141.@xcite more specifically , after some algebra , one gets @xmath142=
\sinh x\cosh(\lambda x)\nonumber\\ & & \times\sin(\lambda y)(\cos y-1)-\sinh ( \lambda x)\cosh x\nonumber\\ & & \times\sin y\left[\cos ( \lambda y)-1\right ] , \label{fw : eq2}\end{aligned}\ ] ] @xmath143^{-1 } , \label{fw : eq3}\ ] ] @xmath144 the inverse of the parameter @xmath139 represents the correlation length @xmath145 . from a practical point of view
it is more convenient to use @xmath139 rather than @xmath108 as a free parameter to construct the fw line . in that way , eq . becomes a transcendental equation that gives @xmath140 as a function of @xmath139 ; once @xmath146 is known , the solution to eq .
gives @xmath108 as a function of @xmath139 ; finally , insertion of @xmath146 and @xmath147 into eq .
provides @xmath148 .
the corresponding values of the pressure are obtained from either eq .
( virial - energy route ) or eq .
( compressibility route ) .
we observe that @xmath13 decreases as @xmath139 decreases , until a _ critical _ value @xmath149 is found in the limit @xmath150 . in that limit ,
@xmath151&= & \sin(\lambda y_c)(\cos y_c-1)-\lambda\sin y_c\nonumber\\ & & \times\left[\cos ( \lambda y_c)-1\right ] , \label{fw : eq5}\end{aligned}\ ] ] @xmath152^{-1 } , \label{fw : eq6}\ ] ] @xmath153 at the critical point @xmath154 and @xmath155 , one has @xmath156 or , equivalently , @xmath157 .
therefore , at this point @xmath43 does not decay for long distances and in fourier space one has @xmath158 and @xmath159 for short wave numbers .
the condition @xmath159 is also satisfied for @xmath160 if @xmath161^{-1}$ ] , in agreement with eq . .
this defines in the @xmath141-@xmath13 plane a spinodal line or locus of points of infinite isothermal compressibility ( within the compressibility route ) .
the spinodal line can not be extended to temperatures larger than the critical value @xmath149 because @xmath162^{-1}\geq\rho_{\max } \sigma$ ] if @xmath163 , where @xmath86 is the maximum density beyond which the hpa is unphysical at a given temperature ( see appendix [ app : appa ] ) .
we further note that the spinodal line only has a lower density ( or vapor - like ) branch , thus hampering the interpretation of @xmath164 as a conventional critical point .
the above features are already suggestive of considering the hpa spinodal line as an artifact of the theory when used in a region of parameter space where the approximation is invalid .
additional support to this view stems from the fact that the hpa keeps predicting a spinodal line and a critical point even in the sw case ( @xmath165 , @xmath166 ) , a clearly incorrect feature . as we shall discuss further below , specialized numerical simulations coupled with a recent analytical study,@xcite strongly support the absence of any phase transition in the present one - dimensional psw model .
vs @xmath108 plane with @xmath81 and @xmath167 ( top panel ) and @xmath111 ( bottom panel ) .
the long dashed curves are the spinodal lines predicted by the hpa . the fw and spinodal liness meet at the critical point ( denoted by a circle ) .
[ fig : fig7 ] ] in fig .
[ fig : fig7 ] we report the comparison between the fw lines , as predicted by the lpa and the hpa , in the @xmath85 vs @xmath97 plane for @xmath81 and two different energy ratios @xmath167 and @xmath111 . the spinodal line predicted by the hpa
is also included .
as said above , the fw and spinodal lines meet at the critical point . while at high temperatures ( above @xmath121 ) there is a remarkable agreement between the two approximations ,
deviations occur at lower temperatures . , but in the @xmath168 vs @xmath108 plane .
note that while in the lpa the three routes to the pressure are not distinguishable one from the other on the graph scale , for the hpa the difference between the virial and the compressibility route is noticeable at low temperatures .
[ fig : fig8 ] ] as in the original work by fisher and widom,@xcite we also report the fw line in the @xmath169 vs @xmath97 plane ( see fig . [ fig : fig8 ] ) , where again we compare the lines as derived from the hpa and lpa schemes . for the energy ratio @xmath167 ( top panel )
we see that the hpa and the lpa give qualitative similar forms of the fw line with a significant deviation at low temperatures , where again the hpa fw line is interrupted at @xmath154 .
note that , while all three standard routes give practically identical results within the lpa , the virial(-energy ) route in the hpa differs from the compressibility one at low temperatures ( more than @xmath102 for @xmath170 ) . for consistency ,
the hpa spinodal line is obtained via the compressibility route only .
similar features occur for the second lower value of the energy ratio , namely @xmath111 ( bottom panel ) .
again , the lpa and hpa lines are qualitatively similar , the three routes in the lpa provide indistinguishable results , and the virial and energy routes in the hpa deviate more than @xmath102 for @xmath170 . the main distinctive feature in this case
@xmath111 is a marked upswing of the tail of the fw line , absent the previous case @xmath167 .
this means that , on increasing penetrability that is , on decreasing @xmath21 the transition from oscillatory ( above the line ) to monotonic ( below the line ) behaviors occurs at a higher pressure and a higher density for a fixed temperature @xmath108 . despite the important differences in the steps followed to derive the lpa and the hpa , it is noteworthy that they agree in the qualitative shape of the fw lines ( even though the hpa predicts a spurious spinodal line ) .
it is then reasonable to expect that the true fw line should interpolate the lpa line at low temperatures with the hpa line at high temperatures .
finally , we now tackle the issue of the existence of a phase transition for the psw one - dimensional model . in view of the hpa results on the seemingly existence of a spinodal line ( and hence of a critical point ) , we here consider the fluid - fluid transition .
as we shall see , our numerics is compatible with the absence of such a transition , thus supporting the view that the above findings of a spinodal line is indeed a consequence of the application of the hpa to a regime where the theory is not valid .
as the fw line always anticipates the critical point , as remarked , we can then look for the existence of a fluid - fluid coexistence line in the region predicted by the interpolation of the lpa and hpa fisher widom lines . we have carried out extensive simulations of the psw fluid using gibbs ensemble monte carlo techniques and employing all standard improvements suggested in the literature.@xcite in order to validate our code , we tested it against the case of the one - dimensional sw potential , where exact analytical predictions for all thermodynamic quantities are available .
we have used up to @xmath171 particles and carefully scanned the temperature range @xmath172 and the density range @xmath173 , as suggested by the fw line ( see fig . [
fig : fig7 ] ) .
we have also considered different values of @xmath21 and @xmath8 for cases giving a significant overlapping probability . in all the cases
we have not found any signature of a fluid - fluid phase separation .
although the _ absence _ of a critical transition is always much more difficult to assess as opposed to its presence , the first scenario is consistent with more than one indication .
the first indication stems from a lattice model counterpart of the one - dimensional psw model .
this is discussed in appendix [ app : appb ] , where the lattice version of the psw model is constructed following standard manipulations with the result that no phase transitions are present for _ finite occupancy_. an additional evidence supporting the absence of any fluid - fluid or freezing transition stems from the very recent exact analytical work alluded to earlier,@xcite that , using the methodology presented in ref .
, concludes that _ no phase transitions _ are present for the psw and ps models in one dimension . obtained from mc simulations for @xmath80 , @xmath174 , @xmath113 , and @xmath175 ( top panel ) and @xmath176 ( bottom panel ) .
[ fig : blob ] ] in our simulations we have also investigated values of @xmath21 and @xmath8 which violate ruelle s stability criterion ( see appendix [ app : ruelle ] ) and thus the one - dimensional psw model is not necessarily stable in the thermodynamic limit . here
the phenomenology turns out to be much more interesting . for sufficiently low temperatures and sufficiently high densities
we observe the formation of a `` blob '' of many - particle clusters ( each made of a large number of overlapping particles ) having a well defined and regular distribution on the axis , and occupying only a portion of the system length . in the `` blob ''
phases the energy per particle grows with the number of particles , thus revealing the absence of a thermodynamic limit .
the transition from a `` normal '' phase to a `` blob '' phase if @xmath177 , where @xmath7 is the integer part of @xmath8 , is illustrated in fig .
[ fig : blob ] .
this figure shows the rdf at @xmath175 ( top panel ) and @xmath176 ( bottom panel ) for @xmath80 , @xmath174 , and @xmath113 . at the lower density
the structure of the psw fluid is qualitatively not much different from that expected if @xmath178 ( compare , for instance , with fig .
[ fig : fig5 ] ) .
however , the structure at the higher density is reminiscent of that of a solid , except that the distribution of particles does not span the whole length ( here @xmath179 ) . instead
, the particles distribute into a few clusters regularly spaced with distance @xmath180 , so that the particles of a given cluster interact attractively with all the particles of the nearest- and next - nearest - neighbor clusters .
note that the number of particles within any given cluster is not necessarily identical .
it is also worth noticing that , in spite of the huge difference in the vertical scales of both panels in fig .
[ fig : blob ] , they are consistent with the condition @xmath181 . the decay of the peaks of @xmath28 is mainly due to the lack of translational invariance , i.e. , the first and last clusters have only one nearest - neighbor cluster , the first , second , next - to - last , and last clusters have only one next - nearest - neighbor cluster , and so on .
we stress that the above phenomenon is specific of bounded potentials , such as psw , and has no counterpart in the hard - core domain .
it is then plausible to expect their appearance even in the corresponding three - dimensional versions of these models where freezing transition ( and phase separation for sw ) are present but could both be hampered by the presence of this clustering phenomenon in the region of parameter space where ruelle s stability criterion is violated .
this would extend the interesting phenomenology already established for the ps case @xcite .
work along these lines on the three - dimensional case is underway and will be reported elsewhere .
in this paper we have completed the study initiated in previous work@xcite on the penetrable - square - well model in one - dimension .
this is a model combining the three main ingredients present in many colloidal - polymer solutions , namely repulsions , attractions , and penetrability .
while the first two are ubiquitous even in simple fluids , the latter is a peculiarity of complex fluids where there exist many examples of colloid - polymer systems which are penetrable ( with some energy cost ) to some extent , and they involve both steric repulsions and short - range attractions .
this model then captures all these crucial features at the simplest level of description within an implicit solvent description . the main new point of this paper was to present an additional and complementary approximation , denoted as the hpa , valid in regimes complementary to those valid for the low - penetrability scheme lpa discussed in ref . .
while the idea behind the lpa was to modify the _ exact _ relations valid for the one - dimensional sw fluid and in fact for any fluid with a hard core and short - range attraction to allow penetrability within some reasonable approximation , the driving force behind the hpa is the fact that for bounded potential the mayer function can be made arbitrarily small by considering sufficiently high temperatures .
as a consequence , only the linear chain diagrams need to be retained at each order in the cluster expansion . as it turns out,@xcite this is tantamount to considering the celebrated debye
hckel theory for charged fluids , and we have here considered the soundness of this approximations at various regimes as compared to specialized mc simulations .
the latter were also compared with other approximations which parallel the entire hierarchy of approximations in the framework of charged fluids , ranging from the most sophisticated non - linear debye
hckel to the simplest mean - spherical approximations .
we have assessed the regime of reliability of these approximations both for thermodynamic and correlation functions by comparison with mc simulations and by internal consistency between different routes to thermodynamics .
next we have also discussed the location of the fw line , separating oscillatory from monotonic behavior in the correlation function , within the hpa and compared with that obtained from the lpa introduced in previous work.@xcite in agreement with previous findings , we find that penetrability enhances the region where correlation functions have a monotonic regime .
the fw derived from the hpa and lpa schemes are found to be in qualitative agreement thus making it possible the drawing of a line interpolating the high- and low - temperature regimes . as a final point
, we investigated the possibility of a fluid - fluid transition .
this possibility arises because the boundness of the potential renders the van hove theorem on the absence of phase transition for one - dimensional model with short - range interactions non - applicable .
in fact , the hpa is seen to predict a critical point where the fw line meets a spinodal line
. however , this prediction takes place in a region of densities and temperatures where the hpa is not reliable .
a careful investigation using both nvt and gibbs ensemble monte carlo techniques akin to those exploited in the investigations of the analog problem for the three - dimensional sw model yields negative results .
these findings are also supported by analytical arguments based on the lattice gas counterpart where the absence of transition can be motivated by the absence of an infinite occupancy of each site , as well as by an exact analytical proof@xcite of a no - go theorem proving the absence of _ any _ phase transition in this model , which is in agreement with , and beautifully complements , our work . in our quest for a possible thermodynamic transition in the one - dimensional psw model
we have explored values of the energy ratio @xmath21 and well width @xmath8 for which the stability of the system in the thermodynamic limit is not guaranteed by ruelle s criterion.@xcite we have found that , as the temperature decreases and/or the density increases , a transition from a normal fluid phase to a peculiar solid - like phase takes place .
the latter phase is characterized by the formation of clusters of overlapping particles occupying a small fraction of the available space and with non - extensive properties . this clustering transition preempts both the fluid - fluid and fluid - solid transitions . in view of the results presented here
, it would be very interesting to discuss the phase diagram of the corresponding three - dimensional psw model .
the phase diagram of the sw model ( @xmath165 ) is indeed well established and includes both a fluid - solid transition present even in the hs counterpart and a fluid - fluid transition line .
the latter is present for any value of @xmath3 and @xmath8 but is stable against freezing only for @xmath182 , the depth of the well being irrelevant .
@xcite the results presented here strongly suggest the importance of the additional parameter @xmath21
. a first interesting issue would be the ruelle instability in three - dimensions . a straightforward extension of the arguments presented in appendix [ app : ruelle ] predicts a guaranteed stability for @xmath183 ( if @xmath4 ) , but the actual onset of the instability can not be assessed through these arguments .
one could expect that for sufficiently high penetrability ( i.e. , @xmath184 and high density ) a phenomenon akin to the `` clustering '' transition found here could be present . in the case
@xmath183 , where the clustering transition is not expected , the interesting point is to assess the influence of the ratio @xmath21 on the location of the fluid - fluid critical point and coexistence line .
all of this opens the possibility of a rich and interesting phase diagram which would complement that already present in a general class of bounded potentials with no attractive tails .
@xcite we note that the high - penetrability regime is indeed the realm of the hpa presented here , which can be obviously extended to three - dimensions .
work on the three - dimensional psw model including the above points and other aspects is underway and will be reported elsewhere .
one of us ( rf ) wishes to thank m. kastner for useful discussions and ongoing collaboration .
he also gratefully acknowledges the support of the nithep of south africa . the support of a prin - cofin 2007b58eab grant is acknowledged .
the research of a.s . was supported by the spanish government through grant no . fis2007 - 60977 , partially financed by feder funds .
am would like to acknowledge the financial support of the msmt of the czech republic under project no .
lc512 and the gaas of the czech republic ( grant no .
iaa400720710 ) .
let us consider the 1d psw model characterized by @xmath21 and @xmath8 .
let us call @xmath7 the integer part of @xmath8 , i.e. , @xmath185 . according to ruelle s criterion , a sufficient condition of thermodynamic stability is @xcite @xmath186 for all configurations @xmath187 , where @xmath188 is a fixed bound
. given the number of particles @xmath189 , we want to obtain the configuration with the minimum potential energy @xmath190 . without loss of generality
we can see any given configuration as a set of @xmath191 clusters @xmath192 , each cluster @xmath193 being made of @xmath194 _ overlapping _ particles ( i.e. , any pair of particles of a given cluster are separated a distance smaller than @xmath39 ) . in ref .
we proved that for a fixed value of @xmath191 the minimum energy corresponds to @xmath195 , all the particles of each cluster being located at the same point and the centers of two adjacent clusters being separated a distance @xmath39 .
therefore , we can restrict ourselves to this class of ordered configurations and use @xmath196 as the variational variable . the repulsive contribution to the potential energy
is @xmath197 to compute the attractive contribution we need to take into account that all the particles of a given cluster interact attractively with the particles of the @xmath198 nearest clusters .
the total number of pairs of interacting clusters are @xmath199+\ell+(\ell-1)+(\ell-2)+\cdots + 1= ( \ell+1)\left(m-1-\ell/2\right)$ ] .
therefore , @xmath200 the total potential energy @xmath201 is @xmath202 we then see that the value that minimizes @xmath203 is @xmath204 .
\label{app4}\ ] ] this value is only meaningful if @xmath205 .
otherwise , @xmath206 . in summary ,
the absolute minimum value of @xmath190 is @xmath207 ^ 2\right\ } , & { \epsilon_r}<{2(\ell+1)\epsilon_a},\\ -(\ell+1)\left(n-\frac{\ell+2}{2}\right)\epsilon_a , & \epsilon_r>2(\ell+1)\epsilon_a .
\end{cases } \label{app5}\ ] ] therefore , if @xmath6 the potential energy is bounded from below by @xmath208 with @xmath209 and thus the system is stable in the thermodynamic limit . on the other hand ,
if @xmath210 there exist configurations that violates ruelle s criterion and so the thermodynamic stability of the system is not guaranteed .
starting from eq . ( [ hpa : eq4 ] ) and for @xmath211 , the fourier transform @xmath134 can be expanded in power series as @xmath212 upon inverse fourier transform one then has @xmath213 where @xmath214 equation ( [ hpa : eq1 ] ) can be rewritten as @xmath215 .
\label{9}\ ] ] therefore , @xmath216}}~.\end{aligned}\ ] ] the origin ( @xmath217 ) is a regular point of @xmath218 and hence of @xmath219 [ but not of each separate term in eq .
( [ appa : eq4 ] ) ] , so we can choose to save the point @xmath217 in eq .
( [ appa : eq3 ] ) either from above or from below .
here we do it from above with the result @xmath220 where the path @xmath221 in the complex @xmath222 plane goes from @xmath223 to @xmath224 and closes itself on the upper plane if @xmath225 and in the lower one if @xmath226 . in eq .
( [ appa : eq3 ] ) we then find @xmath227^{n-1}}{\left(n-1\right ) ! } \theta\left[-r-2p+m-\left(2q - n+m\right)\lambda\right]~.\end{aligned}\ ] ] it is interesting to note that @xmath228 if @xmath229 . thus eq . ( [ appa : eq2 ] ) can be rewritten as @xmath230\}}^\infty\left(\rho\gamma_r\right)^{n-1}w_n\left(r\right)~,\end{aligned}\ ] ] where @xmath231 $ ] is the integer part of @xmath232 . and of the maximum density @xmath233 versus @xmath234 for @xmath106 and @xmath81 .
[ fig : fig0 ] ] the radius of convergence of the series ( [ appa : eq7 ] ) depends on temperature and can be obtained by the same arguments as in the ps case.@xcite from the denominator of eq .
( [ hpa : eq4 ] ) , it follows that the series converges provided that @xmath235 , where @xmath236 here @xmath237 denotes the absolute maximum value of @xmath238 . from eq .
( [ hpa : eq1 ] ) if @xmath239 that maximum corresponds to @xmath217 , i.e. , @xmath240 , and so @xmath241 . on the other hand ,
if @xmath242 the maximum value @xmath237 takes place at @xmath243 and so @xmath244 must be obtained numerically . for sufficiently large values of @xmath245 one
has @xmath246 , so that @xmath247 and this coincides with the maximum physical density ( see below ) .
figure [ fig : fig0 ] shows @xmath244 as a function of @xmath245 for two values of @xmath8 . in the ps limit ( @xmath248 ) one has @xmath249 .
as the strength of the attractive part of the potential ( measured by the product @xmath245 ) increases , the radius of convergence first grows , reaches a maximum , and then decays . except
when @xmath245 is so large that @xmath246 , the maximum @xmath237 corresponds to a negative value of @xmath45 and so the singularity responsible for the radius of convergence is located on the negative real axis .
therefore , @xmath43 is still well defined beyond the radius of convergence , i.e. , for @xmath250 . on the other hand , analogously to the ps case,@xcite the hpa for the psw
fluid becomes unphysical , at a given temperature , for densities larger than a certain value @xmath251 given by the condition @xmath252 where @xmath253 is the absolute maximum value of @xmath45 .
since @xmath254 , it is obvious that @xmath255 . for sufficiently large values of @xmath245 ( actually , for temperatures below the critical value @xmath149 defined in section [ sec : fw ] ) one has @xmath256 , and so @xmath257 . in that case , the line of maximum density becomes a spinodal line , as discussed in section [ sec : fw ] .
figure [ fig : fig0 ] also includes a plot of @xmath258 as a function of @xmath245 for the same two values of @xmath8 . note the kink of the curve @xmath258 for @xmath81 at the critical point ( @xmath259 , @xmath260 ) , so that @xmath257 ( spinodal line ) if @xmath261 .
consider the psw model in one dimension . the grand partition function is @xmath262 where @xmath263 is the thermal de broglie s wavelength and @xmath264 .
\label{appb : eq2}\ ] ] we now discretize the length @xmath265 as a sum of @xmath266 cells of size @xmath267 with occupancy @xmath268 , @xmath269 .
the value of @xmath270 is chosen in the interval @xmath271 , with @xmath272 , so that two particles in the same cell are assumed to interact repulsively and two particles in adjacent cells are assumed to interact attractively .
the integral can then approximated as @xmath273 and the configurational partition function @xmath274 as @xmath275\nonumber\\ & & \times n ! .
\label{appb : eq3}\end{aligned}\ ] ] in eq .
( [ appb : eq3 ] ) the @xmath276 function accounts for the constraint @xmath277 and the factor @xmath278 of the indistinguishability of the particles .
also we have introduced @xmath279 accounting for multi - interactions among cells and @xmath280 which is equal to @xmath2 if @xmath281 and equal to @xmath282 for nearest - neighboring ( nn ) cells . noting that each particle in an @xmath283-cell can either interact with @xmath284 other particles within the same cell or with @xmath285 particles within a nn @xmath286-cell
, we see that @xmath287 substituting into eq .
( [ appb : eq1 ] ) we then find assume a finite length @xmath265 ( and hence a finite number of cells @xmath291 ) with periodic boundary conditions
. the above partition function can then be solved by standard transfer matrix techniques @xmath292 where we have introduced the matrix @xmath293.\end{aligned}\ ] ] if @xmath291 is _ finite _ , one then has in the thermodynamic limit @xmath294 where @xmath295 is the largest eigenvalue of the matrix .
clearly this is analytic and no phase transitions are possible for _ finite occupancy _ @xmath291 .
99 c. marquest and t. a. witten , j. phys .
( paris ) * 50 * , 1267 ( 1989 ) . c. n. likos , phys .
rep . * 348 * , 267 ( 2001 ) .
f. h. stillinger , j. chem . phys . *
65 * , 3968 ( 1976 ) . c. n. likos , m. watzalwek , and h. lwen , phys .
e * 58 * , 3135 ( 1998 ) .
a. santos , r. fantoni , and a. giacometti , phys .
e * 77 * , 051206 ( 2008 ) .
r. fantoni , a. giacometti , al .
malijevsk , and a. santos , j. chem .
phys . * 131 * , 124106 ( 2009 ) .
malijevsk and a. santos , j. chem . phys . * 124 * , 074508 ( 2006 ) .
malijevsk , s. b. yuste , and a. santos , phys .
e * 76 * , 021504 ( 2007 ) . j .-
hansen and i. r. mcdonald _ theory of simple liquids _
( academic press , amsterdam , 2006 ) .
a. santos , r. fantoni , and a. giacometti , j. chem . phys . * 131 * , 181105 ( 2009 ) .
l. van hove , physica ( amsterdam ) * 16 * , 137 ( 1950 ) .
j. a. cuesta and a. sanchez , j. stat .
* 115 * , 869 ( 2004 ) .
a. z. panagiotopoulos , mol . phys .
* 61 * , 813 ( 1987 ) .
a. z. panagiotopoulos , n. quirke , m. stapleton , and d. j. tildesley , mol . phys . * 63 * , 527 ( 1988 )
. b. smit , ph . de smedt , and d. frenkel , mol
. phys . * 68 * , 931 ( 1989 ) .
b. smit and d. frenkel , mol . phys . * 68 * , 951 ( 1989 ) .
r. fantoni and m. kastner , unpublished ( 2010 ) .
m. kastner , rev .
* 80 * , 167 ( 2008 ) .
m. e. fisher and b. widom , j. chem
* 50 * , 3756 ( 1969 ) .
l. acedo and a. santos , phys .
a * 323 * , 427 ( 2004 ) .
a. santos , j. chem .
* 126 * , 116101 ( 2007 ) . c. n. likos , a. lang , m. watzlawek and h. lwen , phys .
e * 63 * , 031206 ( 2001 ) c. n. likos , b. m. mladek , d. gottwald and g. kahl , j. chem .
* 126 * , 224502 ( 2007 ) .
mc results were obtained from a standard nvt ensemble calculation with periodic boundary conditions .
we used @xmath296 particles and @xmath297 mc steps , which were checked to be sufficient for a good accuracy . c. vega , l. f. rull , and s. lago , phys .
e * 51 * , 3146 ( 1995 ) .
fisher and d. ruelle , j. math .
* 7 * , 260 ( 1966 ) .
d. ruelle , _ statistical mechanics : rigorous results _
( benjamin , london , 1969 ) . w. g. t. kranendonk and d. frenkel , mol .
phys . * 64 * , 403 ( 1988 ) .
n. grewe and w. klein , j. math . phys .
* 18 * , 1729 , 1735 ( 1977 ) ; w. klein and n. grewe , j. chem . phys . * 72 * , 5456 ( 1980 ) . | the one - dimensional penetrable - square - well fluid is studied using both analytical tools and specialized monte carlo simulations .
the model consists of a penetrable core characterized by a finite repulsive energy combined with a short - range attractive well .
this is a many - body one - dimensional problem , lacking an exact analytical solution , for which the usual van hove theorem on the absence of phase transition does not apply .
we determine a high - penetrability approximation complementing a similar low - penetrability approximation presented in previous work .
this is shown to be equivalent to the usual debye
hckel theory for simple charged fluids for which the virial and energy routes are identical .
the internal thermodynamic consistency with the compressibility route and the validity of the approximation in describing the radial distribution function is assessed by a comparison against numerical simulations .
the fisher widom line separating the oscillatory and monotonic large - distance behavior of the radial distribution function is computed within the high - penetrability approximation and compared with the opposite regime , thus providing a strong indication of the location of the line in all possible regimes .
the high - penetrability approximation predicts the existence of a critical point and a spinodal line , but this occurs outside the applicability domain of the theory .
we investigate the possibility of a fluid - fluid transition by gibbs ensemble monte carlo techniques , not finding any evidence of such a transition .
additional analytical arguments are given to support this claim .
finally , we find a clustering transition when ruelle s stability criterion is not fulfilled .
the consequences of these findings on the three - dimensional phase diagrams are also discussed . |
when a spin can not fully satisfy energetically all the interactions with its neighbors , it is `` frustrated '' .
this situation occurs when the interactions are in competition with each other or when the lattice geometry does not allow to satisfy all interaction bonds simultaneously as seen for example in the triangular lattice with an antiferromagnetic interaction between the nearest - neighbors .
effects of the frustration in spin systems have been extensively investigated during the last 30 years .
frustrated spin systems are shown to have unusual properties such as large ground state ( gs ) degeneracy , interesting gs symmetries , successive phase transitions with complicated nature , partially disordered phase , reentrance and disorder lines .
frustrated systems still constitute at present a challenge for theoretical , experimental and simulational methods . for recent reviews ,
the reader is referred to ref . .
the nature of the phase transition in strongly frustrated spin systems has been a subject of intensive investigations in the last 20 years . theoretically , these systems are excellent testing grounds for theories and approximations .
many well - established methods such as renormalization group ( rg ) , high- and low - temperature series expansions etc often failed to deal with these systems .
experimentally , data on different frustrated systems show a variety of possibilities : first - order or second - order transitions with unknown critical exponents etc .
( see reviews in ref . ) .
one of the most studied systems is the stacked triangular antiferromagnet ( sta ) : the antiferromagnetic ( af ) interaction between nearest - neighbor ( nn ) spins on the triangular lattice causes a very strong frustration .
it is impossible@xcite to fully satisfy the three af bond interactions on each equilateral triangle .
the gs configuration of both heisenberg and xy models is the well - known 120-degree structure .
the cases of xy ( @xmath0 ) and heisenberg ( @xmath1 ) spins on the sta have been intensively studied since 1987 . for details , see for example the review by delamotte et al@xcite .
let us briefly recall here some main historical developments .
kawamura @xcite has conjectured by a two - loop rg analysis and monte carlo ( mc ) simulations that the transition in @xmath2 and heisenberg models belong each to a new universality class in dimension @xmath3 . since then
there have been many other calculations and simulations with contradictory results .
for example , azaria et al@xcite suggested from a non - linear sigma model that if the transition is not of first order or mean - field tricritical then it should be @xmath4 universality .
numerical simulations@xcite however did not confirm these conjectures .
antonenko et al.@xcite went further in a four - loop rg calculation with a borel resummation technique .
they concluded that the transition is of first order . from 2000 ,
tissier and coworkers @xcite have carried out a nonperturbative rg study of frustrated magnets for any dimension between two and four .
they recovered all known perturbative one - loop results in two and four dimensions as well as for the infinite spin - component number @xmath5 .
they determined @xmath6 for all @xmath7 and found @xmath8 below which the transition is of first order in contradiction with the conjecture of the existence of a new chiral universality class by kawamura.@xcite they explained why theories and simulations have encountered so far many difficulties by the existence of a whole region in the flow diagram in which the flow is slow : the first - order character for @xmath9 is so weak that the transition has a second - order aspect with `` pseudo '' critical exponents .
they calculated these pseudo exponents and found that they coincided with some experimental data .
while this scenario is very coherent , we note that in this nonperturbative rg technique , the real hamiltonian is truncated at the beginning and replaced by an effective one .
however , as will be seen in this paper , the nonperturbative results are well confirmed .
let us recall some results on the xy case .
early mc results on xy sta have been reviewed by loison.@xcite until 2003 , all numerical simulations found a second - order transition with exponents .
a numerical breakthrough has been realized with the results of itakura@xcite who used an improved mc renormalization - group scheme to investigate the renormalization group flow of the effective hamiltonian used in field - theoretical studies for the xy sta .
he found that the xy sta exhibits a clear first - order behavior and there are no chiral fixed points of renormalization - group flow for @xmath10=2 . in 2004 , peles et al@xcite have used a continuous model to study the xy sta by mc simulation .
they found evidence of a first - order transition . in 2006 , kanki et al@xcite , using a microcanonical mc method , have found a first - order signature of the xy sta .
while these recent simulations have demonstrated evidence of first - order transition for the xy sta in agreement with the nonperturbative rg analysis , all of them suffered one or two uncertain aspects : the work of itakura has used a truncated hamiltonian , the work of peles et al has used standard mc methods and the work of kanki et al used a traditional microcanonical mc technique . using a very high - performance technique for weak first - order transitions , the so - called wang - landau flat - histogram method,@xcite we have recently carried out simulations on the xy sta .
we have found clearly a first - order transition in that system confirming results of other authors and putting an end to the controversy which has been lasting for 20 years .
for the heisenberg case , itakura@xcite found , as in the xy case mentioned above , the absence of chiral fixed points of renormalization - group flow .
however , he could not find numerical evidence of the first - order transition .
he predicted that if the transition is of first order for the heisenberg spins , it should occur at much larger lattice sizes which he was not able to perform at that time .
encouraged by the high performance of the wang - landau method , we decided to study the heisenberg case in this work using the full hamiltonian with very large lattice sizes .
as shown below , we find indeed a first - order transition in this case .
the paper is organized as follows .
section ii is devoted to the description of the model and the technical details of the wang - landau ( wl ) methods as applied in the present paper .
section iii shows our results .
concluding remarks are given in section iv .
we consider the stacking of triangular lattices in the @xmath11 direction .
the spins are the classical heisenberg model of magnitude @xmath12 .
the hamiltonian is given by @xmath13 where @xmath14 is the heisenberg spin at the lattice site @xmath15 , @xmath16 indicates the sum over the nn spin pairs @xmath14 and @xmath17 both in the @xmath18 planes and in adjacent planes in the @xmath11 direction . for simplicity
, we suppose the same antiferromagnetic interaction @xmath19 ( @xmath20 ) for both in - plane nn pairs and inter - plane nn ones .
recently , wang and landau@xcite proposed a monte carlo algorithm for classical statistical models .
the algorithm uses a random walk in energy space in order to obtained an accurate estimate for the density of states @xmath21 which is defined as the number of spin configurations for any given @xmath22 .
this method is based on the fact that a flat energy histogram @xmath23 is produced if the probability for the transition to a state of energy @xmath22 is proportional to @xmath24 . at the beginning of the simulation , the density of states ( dos )
is set equal to one for all possible energies , @xmath25 .
we begin a random walk in energy space @xmath26 by choosing a site randomly and flipping its spin with a probability proportional to the inverse of the momentary density of states . in general ,
if @xmath22 and @xmath27 are the energies before and after a spin is flipped , the transition probability from @xmath22 to @xmath27 is @xmath28 .
\label{eq : wlprob}\ ] ] each time an energy level @xmath22 is visited , the dos is modified by a modification factor @xmath29 whether the spin flipped or not , i.e. @xmath30 . at the beginning of
the random walk , the modification factor @xmath31 can be as large as @xmath32 .
a histogram @xmath23 records how often a state of energy @xmath22 is visited .
each time the energy histogram satisfies a certain `` flatness '' criterion , @xmath31 is reduced according to @xmath33 and @xmath23 is reset to zero for all energies .
the reduction process of the modification factor @xmath31 is repeated several times until a final value @xmath34 which close enough to one .
the histogram is considered as flat if @xmath35 for all energies , where @xmath36 is chosen between @xmath37 and @xmath38 and @xmath39 is the average histogram .
the thermodynamic quantities@xcite can be evaluated by @xmath40 , @xmath41 , @xmath42 , and @xmath43 , where @xmath44 is the partition function defined by @xmath45 .
the canonical distribution at any temperature can be calculated simply by @xmath46 . in this work ,
we consider a energy range of interest@xcite @xmath47 .
we divide this energy range to @xmath48 subintervals , the minimum energy of each subinterval is @xmath49 for @xmath50 , and maximum of the subinterval @xmath15 is @xmath51 , where @xmath52 can be chosen large enough for a smooth boundary between two subintervals . the wang - landau algorithm is used to calculate the relative dos of each subinterval @xmath53 with the modification factor @xmath54 and flatness criterion @xmath55 .
we reject the suggested spin flip and do not update @xmath21 and the energy histogram @xmath23 of the current energy level @xmath22 if the spin - flip trial would result in an energy outside the energy segment .
the dos of the whole range is obtained by joining the dos of each subinterval @xmath56 .
we used the system size of @xmath57 where @xmath58 , 84 , 90 , 96 , 108 , 120 and 150 .
periodic boundary conditions are used in the three directions .
@xmath59 is taken as the unit of energy in the following .
the energy histograms for three representative sizes @xmath60 , @xmath61 and @xmath62 shown in figs .
[ fig : stah96pe ] , [ fig : stah120pe ] and [ fig : stah150pe ] , respectively .
as seen , for @xmath60 , the peak is very broad , a signature of the beginning to of a double - maximum structure .
the double peak begins really at @xmath61 .
we note that the distance between the two peaks , i. e. the latent heat , increases with increasing size and reaches @xmath63 for @xmath62 .
this is to be compared with the value @xmath64 for @xmath61 in the xy case.@xcite such a small value of the latent heat in the heisenberg case explains why the first - order character was so difficult to be observed . for increasing sizes ,
the minimum between the peaks will be deepened to separate completely the two peaks .
note that the double - peak structure is a sufficient condition , not a necessary condition , for a first - order transition .
we give here the values of @xmath65 for a few sizes : @xmath66 0.95774 , 0.95768 and 0.957242 for @xmath10=96 , 120 and 150 , respectively .
to explain why standard mc methods without histogram monitoring ( see for example ref . )
fail to see the first order character , let us show in fig .
[ fig : stah3e ] the energy vs @xmath67 obtained by averaging over states obtained by the wl method for @xmath60 , 120 and 150 .
we see here that while the energy histograms show already a signature of double - peak structure at these big sizes , the average energy calculated by using these wl histograms does not show a discontinuity : the averaging over all states erases away the bimodal distribution seen in the energy histogram at the transition temperature .
therefore , care should be taken to avoid such problems due to averaging in mc simulations when studying weak first - order transitions .
figures [ fig : stah3 m ] and [ fig : stah3x ] show the magnetization and the susceptibility for three sizes @xmath60 , 120 and 150 . again here , one does not see with one s eye the discontinuity of the magnetization at the transition even for @xmath62 .
the averaging procedure erases , as for the energy , the detailed structure at the transition . at this stage
it is interesting to make another check of the first - order character : in a first - order transition , the maximum of the susceptibility should scale with the system volume , namely @xmath68 where @xmath7 is the system dimension.@xcite we plot in fig .
[ fig : stahgam ] @xmath69 versus @xmath10 in a @xmath70 scale .
the slope of the straight line is @xmath71 which is nothing but @xmath7 within errors .
this is a very strong signature of a first - order transition .
we have studied in this paper the phase transition in the heisenberg sta by using the flat histogram technique invented by wang and landau .
the method is very efficient because it helps to overcome extremely long transition time between energy valleys in systems with a first - order phase transition .
we found that the transition becomes clearly of first - order only at a very large lattice size confirming the result of a nonperturbative rg calculations using an effective average hamiltonian .
the present work hence puts definitely an end to the long - standing controversial subject on the nature of the phase transition in heisenberg sta . to conclude ,
let us emphasize that for complicated systems like this one , methods well established for simple systems such as ferromagnets may encounter difficulties in dealing with the nature of the phase transition .
such difficulties can be solved only with high - performance mc simulations as the one used here , and a detailed analysis of the flow behavior as suggested by a nonperturbative rg calculation .
one of us ( vtn ) would like to thank nishina memorial foundation for a six - month postdoctoral fellowship .
he is also grateful to prof .
t. ando for hospitality and encouragement during his stay at the tokyo institute of technology .
d. loison and h. t. diep , phys .
b * 50 * , 16453 ( 1994 ) ; t. bhattacharya , a. billoire , r. lacaze and th .
jolicoeur , j. physique i ( france ) * 4 * , 122 ( 1994 ) .
e. h. boubcheur , d. loison and h. t. diep , phys .
b * 54 * , 4165 ( 1996 ) .
a. dobry and h. t. diep , phys .
b * 51 * , 6731 ( 1995 ) ; d. loison and h. t. diep , j. appl . phys . * 76 * , 6350 ( 1994 ) .
g. brown and t.c .
schulhess , j. appl . phys .
* 97 * , 10e303 ( 2005 ) .
b. j. schulz , k. binder , m. mller , and d. p. landau , phys .
e * 67 * , 067102 ( 2003 ) .
a. malakis , s. s. martinos , i. a. hadjiagapiou , n. g. fytas , and p. kalozoumis , phys .
e * 72 * , 066120 ( 2005 ) . | by using the wang - landau flat - histogram monte carlo ( mc ) method for very large lattice sizes never simulated before , we show that the phase transition in the frustrated heisenberg stacked triangular antiferromagnet is of first - order , contrary to results of earlier mc simulations using old - fashioned methods .
our result lends support to the conclusion of a nonperturbative renormalization group performed on an effective hamiltonian .
it puts an end to a 20-year long controversial issue . |
ever since he wrote his large 1907 review of special relativity @xcite for the _ jahrbuch der radioaktivitt und elektronik _
, einstein reflected on how to extend the principle of relativity to non - inertial motions .
his key insight was that such an extension is indeed possible , provided gravitational fields are included in the description .
in fact , the last chapter ( v ) of @xcite , which comprises four ( 17 - 20 ) out of twenty sections , is devoted to this intimate relation between acceleration and gravitation .
the heuristic principle einstein used was his `` quivalenzhypothese '' ( hypothesis of equivalence ) or `` quivalenzprinzip '' ( principle of equivalence ) , which says this : changing the description of a system from an inertial to a non - inertial reference frame is equivalent to not changing the frame at all but adding a _ special _ gravitational field .
this principle is _ heuristic _ in the sense that it allows to deduce the extension of physical laws , the forms of which are assumed to be known in the absence of gravitational fields , to the presence of at least those special gravitational fields that can be `` created '' by mere changes of reference frames .
the idea behind this was , of course , to postulate that the general features found in this fashion remain valid in _ all _ gravitational fields . in the 1907 review einstein used this strategy to find out about the influence gravitational fields have on clocks and general electromagnetic processes .
what he did not attempt back in 1907 was to find an appropriate law for the gravitational field that could replace the poisson equation of newtonian gravity .
this he first attempted in his two `` prague papers '' from 1912 @xcite@xcite for static fields .
the purpose of my contribution here is to point out that the field equation einstein arrived at in the second of these papers is not merely of historical interest .
after 1907 einstein turned away from gravity research for a while , which he resumed in 1911 with a paper @xcite , also from prague , in which he used the `` quivalenzhypothese '' to deduce the equality between gravitational and inertial mass , the gravitational redshift , and the deflection of light by the gravitational field of massive bodies .
as is well known , the latter resulted in half the amount that was later correctly predicted by gr . in the next gravity paper @xcite , the first in 1912 , entitled _
`` lichtgeschwindigkeit und statik des gravitationsfeldes '' _ , einstein pushed further the consequences of his heuristics and began his search for a sufficiently simple differential equation for static gravitational fields .
the strategy was to , first , guess the equation from the form of the special fields `` created '' by non inertial reference frames and , second , generalise it to those gravitational fields sourced by real matter .
note that the gravitational acceleration was to be assumed to be a gradient field ( curl free ) so that the sought - after field equation was for a scalar field , the gravitational potential .
the essential idea in the first 1912 paper is to identify the gravitational potential with @xmath1 , the local velocity of light . has the wrong physical dimension , namely that of a velocity , whereas the a proper gravitational potential should have the dimension of a velocity - squared . ]
einstein s heuristics indicated clearly that special relativity had to be abandoned , in contrast to the attempts by max abraham ( 1875 - 1922 ) , who published a rival theory @xcite@xcite that was superficially based on poincar invariant equations ( but violated special relativity in abandoning the condition that the four - velocities of particles had constant minkowski square ) . in passing i remark that einstein s reply @xcite to abraham , which is his last paper from prague before his return to zrich , contains next to his anticipation of the essential physical hypotheses on which a future theory of gravity could be based ( here i refer to ji bik s contribution to this volume ) , also a concise and very illuminating account of the physical meaning and limitation of the special principle of relativity ,
the essence of which was totally missed by abraham . back to einstein
s first 1912 paper , the equation he came up with was @xmath2 where @xmath3 is the `` universal gravitational constant '' and @xmath4 is the mass density .
the mathematical difference between ( [ eq : firstpragueequation ] ) and the poisson equation in newtonian gravity is that ( [ eq : firstpragueequation ] ) is homogeneous ( even linear ) in the potential @xmath1 .
this means that the source strength of a mass density is weighted by the gravitational potential at its location .
this implies a kind of `` red - shift '' for the active gravitational mass which in turn results in the existence of geometric upper bounds for the latter , as we will discuss in detail below .
homogeneity was einstein s central requirement , which he justified from the interpretation of the gravitational potential as the local velocity of light , which is only determined up to constant rescalings induced from rescalings of the timescale . already in a footnote referring to equation ( [ eq : firstpragueequation ] )
einstein points out that it can not be quite correct , as he is to explain in detail in a follow - up paper @xcite .
this second paper of 1912 is the one i actually wish to focus on in my contribution here .
it appeared in the same issue of the _ annalen der physik _ as the previous one , under the title _
`` zur theorie des statischen gravitationsfeldes '' _ ( on the theory of the static gravitational field ) . in it
einstein once more investigates how the gravitational field influences electromagnetic and thermodynamic processes according to what he now continues to call the _ quivalenzprinzip _ , and derives from it the equality of inertial and gravitational mass .
after that he returns to the equation for the static gravitational field and considers the gravitational force - density @xmath5 , acting on ponderable matter of mass density @xmath4 , which is given by ( einstein writes @xmath6 instead of our @xmath4 ) @xmath7 einstein observes that the space integral of @xmath5 does not necessarily vanish on account of ( [ eq : firstpragueequation ] ) , in violation of the principle that _ actio _ equals _ reactio_. terrible consequences , like self - acceleration , have to be envisaged .
he then comes up with the following non - linear but still homogeneous modification of ( [ eq : firstpragueequation ] ) ( again einstein writes @xmath6 instead of @xmath4 ) : @xmath8 in the rest of this paper we will show how to arrive at this equation from a different direction and discuss some of its interesting properties as well as its relation to the description of static gravitational fields in gr .
the following considerations are based on @xcite .
we start from ordinary newtonian gravity , where the gravitational field is described by a scalar function @xmath9 whose physical dimension is that of a velocity - squared .
it obeys @xmath10 the force per unit volume that the gravitational field exerts onto a distribution of matter with density @xmath4 is @xmath11 this we apply to the force that the gravitational field exerts onto its own source during a real - time process of redistribution .
this we envisage as actively transporting each mass element along the flow line of a vector field @xmath12 .
to first order , the change @xmath13 that @xmath4 suffers in time @xmath14 is given by @xmath15 where @xmath16 and @xmath17 is the lie derivative with respect to @xmath18 .
we assume the support @xmath19 to be compact . in general , this redistribution costs energy .
the work we have to invest for redistribution is , to first order , just given by @xmath20 where we used ( [ eq : densitychange ] ) in the last step and where we did not write out the lebesgue measure @xmath21 to which all integrals refer .
note that in order to obtain ( [ eq : investedworkgeneral ] ) we did not make use the field equation .
equation ( [ eq : investedworkgeneral ] ) is generally valid whenever the force - density relates to the potential and the mass density as in ( [ eq : newtonianforcedensity ] ) .
now we make use of the field equation ( [ eq : newtonianfieldeq ] ) .
we assume the redistribution - process to be adiabatic , that is , we assume the instantaneous validity of the field equation at each point in time throughout the process .
this implies @xmath22 hence , using ( [ eq : investedworkgeneral ] ) , the work invested in the process of redistribution is ( to first order ) @xmath23 if the infinitely dispersed state of matter is assigned the energy - value zero , then the expression in curly brackets is the total work invested in bringing the infinitely dispersed state to that described by the distribution @xmath4 .
this work must be stored somewhere as energy . like in electro - statics and -dynamics ,
we take a further logical step and assume this energy to be spatially distributed in the field according to the integrand .
this leads to the following expression for the energy density of the static gravitational field @xmath24 all this is familiar from newtonian gravity .
but now we go beyond newtonian gravity and require the validity of the following * principle .
* _ all energies , including that of the gravitational field itself , shall gravitate according to @xmath0 . _
this principle implies that if we invest an amount of work @xmath25 to a system its ( active ) gravitational mass will increase by @xmath26 .
now , the ( active ) gravitational mass @xmath27 is defined by the flux of the gravitational field to spatial infinity ( i.e. through spatial spheres as their radii tend to infinity ) : @xmath28 hence , making use of the generally valid equation ( [ eq : investedworkgeneral ] ) , the principle that @xmath29 takes the form @xmath30 this functional equation relates @xmath9 and @xmath4 , over and above the restriction imposed on their relation by the field equation .
however , the latter may - and generally will - be inconsistent with this additional equation .
for example , the newtonian field equation ( [ eq : newtonianfieldeq ] ) is easily seen to manifestly violate ( [ eq : theprinciple ] ) , for the right - hand side then becomes just the integral over @xmath31 , which always vanishes on account of ( [ eq : densitychange ] ) ( or the obvious remark that the redistribution clearly does not change the total mass ) , whereas the left hand side will generally be non - zero . the task must therefore be to find field equation(s ) consistent with ( [ eq : theprinciple ] ) .
our main result in that direction is that the unique generalisation of ( [ eq : newtonianfieldeq ] ) which satisfies ( [ eq : theprinciple ] ) is just ( [ eq : secondpragueequation ] ) , i.e. the field equation from einstein s second 1912 paper .
let us see how this comes about .
a first guess for a consistent modification of ( [ eq : newtonianfieldeq ] ) is to simply add @xmath32 to the source @xmath4 : @xmath33 but this can not be the final answer because this change of the field equation also brings about a change in the expression for the self - energy of the gravitational field .
that is , the term in the bracket on the right - hand side is not the total energy according to _ this _ equation , but according to the original equation ( [ eq : newtonianfieldeq ] ) . in other words :
equation ( [ eq : improvedng - firstiteration ] ) still lacks _ _ self-__consistency .
this can be corrected for by iterating this procedure , i.e. , determining the field s energy density according to ( [ eq : improvedng - firstiteration ] ) and correcting the right - hand side of ( [ eq : improvedng - firstiteration ] ) accordingly .
again we have changed the equation , and this goes on ad infinitum .
but the procedure converges to a unique field equation , similarly to the convergence of the `` noether - procedure '' that leads from the poincar invariant pauli - fierz theory of spin-2 mass-0 fields in flat minkowski space to gr @xcite@xcite@xcite . in our toy model
the convergence of this procedure is not difficult to see .
we start from the definition ( [ eq : defgravmass ] ) and calculate its variation @xmath34 assuming the validity of ( [ eq : improvedng - firstiteration ] ) . from what we said above we know already that this not yet going to satisfy ( [ eq : theprinciple ] ) .
but we will see that from this calculation we can read off the right redefinitions .
we start by varying ( [ eq : defgravmass ] ) : @xmath35 we replace @xmath36 with the variation of the right - hand side of ( [ eq : improvedng - firstiteration ] ) .
partial integration of the non - liner part gives us a surface term whose integrand is @xmath37 and hence vanishes .
the remaining equation is @xmath38 playing the same trick ( of replacing @xmath36 with the variation of the right - hand side of ( [ eq : improvedng - firstiteration ] ) and partial integration , so as to collect all derivatives on @xmath39 ) again and again , we arrive after @xmath40 steps at @xmath41 as @xmath9 is bounded for a regular matter distribution , and the spatial integral over @xmath42 is just @xmath43 , the last term tends to zero for @xmath44 .
hence @xmath45 this _ is _ of the desired form ( [ eq : theprinciple ] ) required by the principle , provided we redefine the gravitational potential to be @xmath46 rather than @xmath9 , where @xmath47 saying that @xmath46 rather than @xmath9 is the right gravitational potential means that the force density is not given by ( [ eq : newtonianforcedensity ] ) , but rather by @xmath48 as we have made use of equation ( [ eq : improvedng - firstiteration ] ) in order to derive ( [ eq : laststep ] ) , we must make sure to keep _ that _ equation , just re - expressed in terms of @xmath46 .
this leads to @xmath49\,,\ ] ] which is precisely einsteins improved `` prague equation '' ( [ eq : secondpragueequation ] ) with @xmath50 .
note from ( [ eq : newpotential ] ) that the asymptotic condition @xmath51 translates to @xmath52 .
note also that for @xmath53 the @xmath54-parts of @xmath55 and @xmath56 coincide , so that in the expressions ( [ eq : defgravmass ] ) for @xmath27 we may just replace @xmath9 with @xmath46 : @xmath57 the principle now takes the form ( [ eq : theprinciple ] ) with @xmath9 replaced by @xmath46 .
it is straightforward to show by direct calculation that ( [ eq : theprinciple ] ) is indeed a consequence of ( [ eq : selfconsistenteq ] ) , as it must be .
it also follows from ( [ eq : selfconsistenteq ] ) that the force density ( [ eq : newforcedensity ] ) is the divergence of a symmetric tensor : [ eq : divsymmtensor ] @xmath58 where @xmath59 \right\}\,.\ ] ] this implies the validity of the principle that actio equals reactio that einstein demanded .
_ this _ was einstein s rationale for letting ( [ eq : secondpragueequation ] ) replace ( [ eq : firstpragueequation ] ) .
finally we mention that ( [ eq : selfconsistenteq ] ) may be linearised if written in terms of the square - root of @xmath46 : @xmath60 one gets @xmath61 this helps in finding explicit solutions to ( [ eq : selfconsistenteq ] ) .
note that @xmath62 is dimensionless .
in this section we discuss some properties of spherically symmetric solutions to ( [ eq : selfconsistenteqlinear ] ) for spherically symmetric mass distributions @xmath4 of compact support . in the following
we will simply refer to the object described by such a mass distribution as `` star '' . in terms of @xmath63 equation ( [ eq : selfconsistenteqlinear ] )
is equivalent to @xmath64 the support of @xmath4 is a closed ball of radius @xmath65 , called the star s radius . for @xmath66 we shall assume @xmath67 ( weak energy condition ) .
we seek solutions which correspond to everywhere positive and regular @xmath62 and hence everywhere positive and regular @xmath46 .
in particular @xmath68 and @xmath69 must be finite
. for @xmath70 equation ( [ eq : selfconsistenteqlinear - chi ] ) implies @xmath71 , the solution to which is @xmath72 where @xmath73 denotes the gravitational radius @xmath74 @xmath73 comes in because of ( [ eq : defgravmassnew ] ) , which fixes one of the two integration constants , the other being fixed by @xmath75 .
let @xmath76 denote the solution in the interior of the star .
continuity and differentiability at @xmath77 gives @xmath78 and @xmath79 .
we observe that @xmath80 . for suppose @xmath81 , then ( [ eq : selfconsistenteqlinear - chi ] ) and the weak energy condition imply @xmath82 . but
this implies that for @xmath83 $ ] the curve @xmath84 lies below the straight line @xmath85 and assumes a value less than @xmath86 at @xmath87 , in contradiction to the finiteness of @xmath69 which implies @xmath88 .
hence we have the gravitational radius of a spherically symmetric star is universally bound by its ( geometric ) radius , @xmath89 .
equivalently expressed in terms of @xmath27 we may say that the gravitational mass is universally bound above by @xmath90 this may be seen in analogy to buchdahl s inequality in gr @xcite , which , using the isotropic ( rather than schwarzschild ) radial coordinate , would differ from ( [ eq : univmassupperbound ] ) only by an additional factor of @xmath91 on the right - hand side .
the buchdahl bound is optimal , being saturated by the interior schwarzschild solution for a homogeneous star .
so let us here , too , specialise to a homogeneous star , @xmath92 where @xmath93 is called the bare mass ( integral over @xmath4 ) .
it is convenient to introduce the radii corresponding to bare and gravitational masses , as well as their ratio to the star s
radius @xmath65 : [ eq : defradii ] @xmath94 we also introduce the inverse length @xmath95 so that for ( [ eq : selfconsistenteqlinear - chi ] ) just reads @xmath96 . from this
the interior solution is easily obtained .
if written in terms of @xmath62 it reads @xmath97 as a result of the matching to the exterior solution given in ( [ eq : psilarger ] ) , @xmath73 is determined by @xmath65 and @xmath98 , i.e. @xmath65 and @xmath99 . in terms of @xmath100 and
@xmath101 this relation takes the simple form @xmath102 which convex - monotonically maps @xmath103 onto @xmath104 . the fact that @xmath105 for all @xmath100 is just the statement of the theorem applied to the homogeneous case . if @xmath106 we have @xmath107 , which for @xmath108 reads @xmath109 we note that @xmath110 is just the newtonian binding energy of a homogeneous star
in view of our principle it makes good sense that to first order just this amount is subtracted from the bare mass in order to obtain the active gravitational mass . in newtonian gravity this negative amount is just identified with the field s self - energy , but here the interpretation is different : the two terms that act as source for the gravitational field in ( [ eq : selfconsistenteq ] ) are the matter part , which is proportional to @xmath4 but diminished by @xmath46 , and the field s own part , which is proportional to @xmath111 and positive definite !
their contributions are , respectively , @xmath112 hence even though the total energy is decreased due to binding , the gravitational field s self energy _
increases _ by the same amount . twice that amount is gained from the fact that the matter - energy is `` red - shifted '' by being multiplied with @xmath46 , so energy is conserved ( of course ) .
two more consequences , which are related , are noteworthy : * unlike in newtonian theory , objects with non - zero gravitational mass can not be modelled by point sources . in the spherically symmetric case
this is an immediate consequence of ( [ eq : univmassupperbound ] ) , which implies @xmath113 for @xmath114 .
hence there are no @xmath115-like masses . *
unlike in newtonian gravity , unlimited compression of matter does not lead to unlimited energy release . consider a sequence of homogeneous ( just for simplicity ) stars of fixed bare mass @xmath93 and variable radius @xmath65 , then the gravitational mass @xmath27 as function of @xmath116
is given by @xmath117 the function in curly brackets is @xmath118 . ]
is a strictly monotonically decreasing function @xmath119\mapsto[1,0]$ ] .
this shows that for infinitely dispersed matter , where @xmath120 and hence @xmath121 , we have @xmath122 , as expected , and that for infinite compression @xmath123 . as the gained energy at stage @xmath100 is @xmath124 , we can at most gain @xmath125 .
finally i wish to briefly comment on the relation of equation ( [ eq : secondpragueequation ] ) or ( [ eq : selfconsistenteq ] ) to gr .
since einstein s 1912 theory was only meant to be valid for static situations , i will restrict attention to static spacetimes @xmath126 .
hence i assume the existence of a timelike and hypersurface orthogonal killing field @xmath127 .
my signature convention shall be `` mostly plus '' , i.e. @xmath128 .
we choose adapted coordinates @xmath129 , @xmath130 , where the level sets of @xmath131 are the integral manifolds of the foliation defined by @xmath127 and @xmath132 .
we can then write the metric in a form in which the coefficients do not depend on @xmath131 ( called `` time '' ) , @xmath133 clearly @xmath134 . from now on , all symbols with hats on refer to the spatial geometry , like the spatial metric @xmath135 .
the @xmath131-component of the geodesic equation is equivalent to @xmath136 , where an overdot refers to the derivative with respect to an affine parameter .
this equation allows us to eliminate the affine parameter in favour of @xmath131 in the spatial components of the geodesic equation . if we set ) which we need and to which we return below . ]
they read @xmath138\,,\ ] ] where the @xmath139 are the christoffel coefficients for @xmath135 , and @xmath140 .
this should be compared with ( [ eq : newforcedensity ] ) together with newton s second law , which give @xmath141 .
as we did not attempt to include special relativistic effects in connection with high velocities , we should consistently neglect terms @xmath142 in ( [ eq : geodesiceqspatial ] ) .
this results in dropping the rightmost term .
the rest has the pseudo - newtonian form in arbitrary ( not just inertial ) spatial coordinates . a non - zero spatial curvature would , of course , be a new feature not yet considered .
the curvature and ricci tensors for the metric ( [ eq : staticmetric ] ) are readily computed , most easily by using cartan s structure equations : @xmath143 here @xmath144 is the unit timelike vector characterising the static reference frame , @xmath145 is the levi - civita covariant derivative with respect to @xmath135 , and @xmath146 is the corresponding laplacian . using this in einstein s equations @xmath147 for pressureless ( we neglect the pressure since it enters multiplied with @xmath148 ) dust at rest and of mass - density @xmath4 in the static frame , i.e. @xmath149 we get we note that , apart from the space curvature , ( [ eq : einsteinsequationshere - a ] ) is almost but not quite identical to ( [ eq : selfconsistenteqlinear ] ) .
they differ by a factor of 2 ! rewriting ( [ eq : einsteinsequationshere - a ] ) in terms of @xmath46 according to ( [ eq : phipsirelation ] ) , we get @xmath151\,.\ ] ] this differs from ( [ eq : selfconsistenteq ] ) by the same factor of @xmath152 ( i.e. , @xmath153 ) .
note that we can not simply remove this factor by rescaling @xmath62 and @xmath46 , as the equations are homogeneous in these fields .
note also that the overall scale of @xmath46 is fixed by ( [ eq : geodesiceqspatial ] ) : it is the gradient of @xmath46 , and not a multiple thereof , which gives the acceleration .
but then there is another factor of 2 in difference to our earlier discussion : if the metric ( [ eq : staticmetric ] ) is to approach the minkowski metric far away from the source , then @xmath62 should tend to one and hence @xmath46 should asymptotically approach @xmath154 according to ( [ eq : phipsirelation ] ) . in ( [ eq : selfconsistenteq ] ) , however , @xmath46 should asymptotically approach @xmath155 , i.e. twice that value . this additional factor of 2 ensures that both theories have the same newtonian limit . indeed ,
if we expand the gravitational potential @xmath46 of an isolated object in a power series in @xmath156 , this implies that the linear terms of both theories coincide .
however , the quadratic terms in gr are twice as large as in our previous theory based on ( [ eq : newforcedensity ] ) and ( [ eq : selfconsistenteq ] ) .
this is not quite unexpected if we take into account that in gr we also have the space curvature that will modify the fields and geodesics in post newtonian approximations .
we note that the spatial einstein equations ( [ eq : einsteinsequationshere - b ] ) prevent space from being flat .
for example , taking their trace and using ( [ eq : einsteinsequationshere - a ] ) shows that the scalar curvature of space is , in fact , proportional to the mass density .
finally we show that the total gravitational mass in gr is just given by the same formula ( [ eq : defgravmassnew ] ) , where @xmath46 is now that used here in the gr context . to see this we recall that for spatially asymptotically flat
spacetimes the overall mass ( measured at spatial infinity ) is given by the adm - mass .
moreover , for spatially asymptotically flat spacetimes which are stationary and satisfy einstein s equations with sources of spatially compact support , the adm mass is given by the komar integral ( this is , e.g. , proven in theorem 4.13 of @xcite ) .
hence we have @xmath157 here @xmath132 , and @xmath158 is the corresponding 1-form .
the star , @xmath159 , denotes the hodge - duality map . using ( [ eq : phipsirelation ] ) and asymptotic flatness
it is now straightforward to show that the right hand side of ( [ eq : komarmass ] ) can indeed be written in the form of the middle term in ( [ eq : defgravmassnew ] ) .
this term only depends on @xmath46 at infinity , i.e. the newtonian limit , and hence gives a value independent of the factor-2 discrepancy discussed above . in this sense
we may say that the active gravitational mass @xmath27 defined earlier corresponds to @xmath160 in the gr context .
* acknowledgements .
* i sincerely thank the organisers and in particular ji bik for inviting me to the most stimulating and beautiful conference _ relativity and gravitation 100 years after einstein in prague . _ | i reconsider einstein s 1912 `` prague - theory '' of static gravity based on a scalar field obeying a non - linear field equation .
i point out that this equation follows from the self - consistent implementation of the principle that _ all _ energies source the gravitational field according to @xmath0 .
this makes it an interesting toy - model for the `` flat - space approach '' to general relativity ( gr ) , as pioneered by kraichnan and later feynman .
solutions modelling stars show features familiar from gr , e.g. , buchdahl - like inequalities .
the relation to full gr is also discussed .
this lends this toy theory also some pedagogical significance .
this paper is based on a talk delivered at the conference _ relativity and gravitation 100 years after einstein in prague _ , held in prague 25.-29 .
june 2012 . |
the study of the structure of finite dimensional algebras has gained importance in recent times for it may be possible to find presumably new examples of subfactors of a hyper finite @xmath4-factor along the lines of @xcite .
the partition algebras are introduced by v. jones in @xcite and martin in @xcite and they are studied intensively by martin in @xcite .
the cellularity of many subalgebras of partition algebras like tanabe algebras , party algebras , algebra of @xmath3-relations and signed partition algebras are studied in detail in @xcite , @xcite , @xcite and @xcite respectively .
the question of semisimplicity of these algebras reduces to the nondegeneracy of gram matrices of such algebras . in this connection ,
the gram matrices of partition algebras , algebra of @xmath3-relations and signed partition algebras are studied in @xcite which are realized as direct sum of block sub matrices .
the diagonal entry of the block submatrix is a product of @xmath5 quadratic polynomials and @xmath6 linear polynomials in the case of algebra of @xmath3-relations and signed partition algebras and it is a product @xmath7 linear polynomials in the case of partition algebras . in this paper , we introduce symmetric diagram matrices @xmath0 of size @xmath1 whose entries are @xmath2 based on the diagrams .
we compute the eigenvalues of symmetric diagram matrices using elementary row and column operations by induction . as a byproduct , we compute the eigenvalues of gram matrices of a larger class of diagram algebras like the signed partition algebras , partition algebras and algebra of @xmath3 relations where the gram matrix of any diagram algebra is a direct product of symmetric diagram matrices .
* + * we recall the definitions in @xcite required in this paper .
let @xmath8 be any set and let @xmath9 denote the set of all equivalence relations on @xmath10 for @xmath11 let @xmath12 let @xmath13 and @xmath14 be the set of all partitions of @xmath15 or equivalence relation on @xmath16 for @xmath17 the elements of @xmath18 are called as connected components .
any @xmath19 can be represented as a simple graph on two rows of @xmath20-vertices , one above the other with @xmath20 vertices in the top row labeled @xmath21 left to right and @xmath20 vertices in the bottom row labeled @xmath22 left to right with vertex @xmath23 and vertex @xmath24 connected by a path if @xmath23 and @xmath24 are in the same block of the set partition @xmath25 the graph representing @xmath18 is called @xmath20-partition diagram and it is not unique .
two @xmath20-partition diagrams are said to be equivalent if they give rise to the same set partition of @xmath26-vertices
. a connected component of @xmath18 containing an element of @xmath27 and element of @xmath28 is called _ through class _ and a connected component containing an element of either @xmath27 or @xmath28 is called _
horizontal edge_. number of through classes in @xmath18 is called _ propagating number _ and it is denoted by @xmath29 we shall define multiplication of two @xmath20-partition diagrams @xmath30 and @xmath31 as follows : * place @xmath30 above @xmath31 and identify the bottom dots of @xmath30 with the top dots of @xmath32 * @xmath33 is the resultant diagram obtained by using only the top row of @xmath30 and bottom row of @xmath34 replace each connected component which lives entirely in the middle row by the variable @xmath35 .
i.e. , @xmath36 where @xmath37 is the number of connected components that lie entirely in the middle row .
this product is associative and is independent of the graph we choose to represent the @xmath20-partition diagram .
let @xmath38 be the field of rational functions with complex coefficients in the variable @xmath39 the partition algebra @xmath40 is defined to be the @xmath38-span of the @xmath20-partition diagrams , which is an associative algebra with identity @xmath41 where by convention @xmath42 for @xmath43 and @xmath44 the following are the generators of the partition algebras .
the above generators satisfy the relations given in theorem 1.11 of @xcite .
[ d2.1 ] ( @xcite ) let the group @xmath3 act on the set @xmath10 then the action of @xmath3 on @xmath8 can be extended to an action of @xmath3 on @xmath45 where @xmath9 denote the set of all equivalence relations on @xmath8 , given by @xmath46 where @xmath47 and @xmath48 ( it is easy to see that the relation @xmath49 is again an equivalence relation ) .
an equivalence relation @xmath18 on @xmath8 is said to be a @xmath3-stable equivalence relation if @xmath50 in @xmath18 implies that @xmath51 in @xmath18 for all @xmath52 in @xmath53 we denote @xmath54 for the set @xmath55 we shall only consider the case when @xmath3 acts freely on @xmath8 ; let @xmath56 and the action is defined by @xmath57 for all @xmath58 let @xmath59 be the set of all @xmath3-stable equivalence relations on @xmath10 [ n2.2](@xcite ) @xmath59 denotes the set of all @xmath3-stable equivalence relation on @xmath60 each @xmath61 can be represented as a simple graph on a row of @xmath26 vertices . *
the vertices @xmath62 are arranged from left to right in a single row . *
if @xmath63 then @xmath64 is joined by a line for all @xmath65 we say that the two graphs are equivalent if they give rise to the same set partition of the @xmath26 vertices @xmath66 we may regard each element @xmath18 in @xmath67 as a @xmath26-partition diagram by arranging the @xmath68 vertices @xmath69 of @xmath18 in two rows in such a way that @xmath70 is in the top(bottom ) row of @xmath18 if @xmath71 for all @xmath72 and if @xmath73 then @xmath64 is joined by a line for all @xmath65 the diagrams @xmath74 and @xmath75 are obtained from the diagram @xmath18 by restricting the vertex set to @xmath76 and @xmath77 respectively .
the diagrams @xmath74 and @xmath75 are also @xmath3-stable equivalence relation and @xmath78 [ d2.3](@xcite ) let @xmath79 then the equation @xmath80 defines an equivalence relation on @xmath16 [ r2.4 ] ( @xcite ) for @xmath81 and for every @xmath3-stable equivalence class or a connected component @xmath82 in @xmath83 there exists a unique subgroup denoted by @xmath84 where 1 .
@xmath85 if @xmath86 is called an @xmath87-class or @xmath87-component and the @xmath87-component @xmath82 will always occur as a pair and 2 .
@xmath88 if @xmath89 is called a @xmath3-class or @xmath3-component and the number of vertices in the @xmath3-component @xmath82 will always be even .
[ p2.5 ] ( @xcite ) the linear span of @xmath67 is a subalgebra of @xmath90 .
we denote this subalgebra by @xmath91 called the algebra of @xmath3-relations .
[ d2.6](@xcite ) for @xmath92 define @xmath93 as follows : @xmath94 i.e. , @xmath18 has @xmath95 number of pairs of @xmath87-through classes and @xmath96 number of @xmath3-through classes .
it is clear that @xmath97 [ d2.7 ] ( @xcite , * definition 3.1.1 * ) let the signed partition algebra @xmath98 be the subalgebra of @xmath90 generated by the subalgebra of the signed partition algebra generated by @xmath99 is isomorphic on to the partition algebra @xmath100 [ d2.8](@xcite , * definition 3.1.2 * ) let @xmath81 . for @xmath101 @xmath102
where 1 .
@xmath103 is a through class of @xmath83 such that @xmath104 2 .
@xmath105 is a through class of @xmath83 such that @xmath106 3 .
@xmath107 is the number of @xmath87 horizontal edges @xmath82 in the top(bottom ) row of @xmath83 such that @xmath85 and @xmath108 4 .
@xmath109 is the number of @xmath3 horizontal edges @xmath82 in the top(bottom ) row of @xmath83 such that @xmath110 5 .
@xmath111 the linear span of @xmath112 is nothing but the signed partition algebra @xmath113 [ r2.9 ] the algebra generated by @xmath83 where @xmath114 is isomorphic to the partition algebra @xmath115 also , let @xmath116 be the set of all @xmath20-partition diagrams @xmath83 in @xmath40 such that @xmath117 where @xmath118 which is contained in @xmath119 [ d2.10 ] * + * 1 .
let @xmath81 with @xmath95 number of pairs of @xmath87-through classes , @xmath96 number of @xmath3-through classes , @xmath5 number of pairs of @xmath87-horizontal edges and @xmath6 number of @xmath3-horizontal edges .
+ let the number of vertices in the top row of @xmath95 number of pairs of @xmath120-through classes(@xmath96 number of @xmath3-through classes ) be denoted by @xmath121 and let the number of vertices in the top row of @xmath5 number of pairs of @xmath87-horizontal edges ( @xmath6 number of horizontal edges ) be denoted by @xmath122 such that + @xmath123 + where @xmath124 with @xmath125 and @xmath74 is obtained from @xmath18 by restricting the vertex set to @xmath66 2 .
let @xmath19 with @xmath126 number of through classes and @xmath7 number of horizontal edges .
+ let the number of vertices in the top row of @xmath126 number of through classes be denoted by @xmath127 and let the number of vertices in the top row of @xmath7 number of horizontal edges be denoted by @xmath128 such that + @xmath129 + where @xmath130 with @xmath131 and @xmath74 is obtained from @xmath18 by restricting the vertex set to @xmath55 [ d2.11]*(@xcite , definition 3.1 ) * define , 1 .
@xmath132 ^ 1 \left [ 2\lambda_2 \right]^2 \left [ \lambda_3 ^ 2 \right]^3 \left [ 2 \lambda_4 \right]^4 \big| \lambda_1 \vdash k_1 , \lambda_2 \vdash k_2 , \lambda_3 \vdash k_3 , \lambda_4 \vdash k_4 \text { with } \lambda_1
\in \mathbb{p}(k_1 , s_1 ) , $ ] + @xmath133 + where @xmath134 + and @xmath135 2 .
@xmath136 ^ 1 [ \lambda_2]^2 \ | \
\lambda_1 \in \mathbb{p}(k_1 , s ) , \lambda_2 \in \mathbb{p}(k_2 , r ) \text { such that } k_1 + k_2 = k\}.$ ] [ n2.12 ] * ( @xcite , notation 3.5 ) * 1 . for
@xmath137 @xmath138 and @xmath139 put @xmath140 + where @xmath141 and @xmath75 are the same , @xmath142 has @xmath5 number of pairs of @xmath87-horizontal edges and @xmath6 number of @xmath3-horizontal edges , @xmath143 denotes the top row of @xmath144 is a subset of @xmath18 such that @xmath145 + also , @xmath146 and @xmath147 2 . for @xmath148 @xmath138 and @xmath149 and if @xmath150 then @xmath151 or @xmath152 + put , @xmath153 + where @xmath154 and @xmath155 are the same , @xmath156 has @xmath5 number of pairs of @xmath87-horizontal edges and @xmath6 number of @xmath3-horizontal edges , @xmath157 is the top row of @xmath158 is a subset of @xmath159 such that @xmath160 + also , @xmath161 and @xmath162 3
. for @xmath163 put @xmath164 where + @xmath165 + @xmath166 + also , @xmath167 [ d2.13 ] * ( @xcite , definition 3.6 ) * 1 .
the diagrams in @xmath168 are indexed as follows : + @xmath169 + @xmath170 * if @xmath171 * if @xmath172 and @xmath173 * if @xmath174 and @xmath175 then it can be indexed arbitrarily .
+ where @xmath5 is the number of pairs of @xmath87-horizontal edges in @xmath176 is the number of pairs of @xmath87-horizontal edges in @xmath177 is the number of @xmath3-horizontal edges in @xmath178 and @xmath179 is the number of @xmath3-horizontal edges in @xmath180 2 .
the diagrams in @xmath181 are indexed as follows : + @xmath182 + @xmath170 1 . if @xmath183 , 2 .
if @xmath184 then it can be indexed arbitrarily + where @xmath185 is the number of horizontal edges in @xmath186 [ d2.14 ] * ( @xcite , definition 3.7 ) * 1 .
* gram matrices of the algebra of @xmath3-relations : * + for @xmath187 define gram matrices of the algebra of @xmath3-relations @xmath188 as follows : + @xmath189 + where @xmath190 denotes the block matrix whose entries are @xmath191 with + @xmath192 + @xmath193 where @xmath194 denotes the number of connected components in @xmath195 excluding the union of all the connected components of @xmath196 and @xmath197 or equivalently , @xmath198 is the number of loops which lie in the middle row when @xmath199 is multiplied with @xmath200 , @xmath201 and @xmath202 respectively .
* gram matrices of signed partition algebra : * + for @xmath203 define gram matrix of the signed partition algebra @xmath204 as follows : + @xmath205 + where @xmath206 denotes the block matrix whose entries are @xmath191 with + @xmath207 + @xmath208 where @xmath209 denotes the number of connected components in @xmath210 excluding the union of all the connected components of @xmath211 and @xmath212 or equivalently , @xmath213 is the number of loops which lie in the middle row when @xmath214 is multiplied with @xmath215 , @xmath216 and @xmath217 respectively .
* gram matrices of partition algebra : * + for @xmath218 define gram matrix of the partition algebra @xmath219 as follows : + @xmath220 + where @xmath221 denotes the block matrix whose entries are @xmath191 with + @xmath222 + @xmath223 where @xmath224 denotes the number of connected components which lie in the middle row while multiplying @xmath225 with @xmath226 @xmath227 and @xmath228 respectively .
[ n2.15 ] * ( @xcite , notation 3.21 ) * 1 .
let @xmath229 such that @xmath230 so that the @xmath231-entry of the block matrix @xmath232 is zero and @xmath233 + put @xmath234 and @xmath235 where @xmath236 is the sub diagram of @xmath237 and every @xmath238 through class of @xmath239 is replaced by a @xmath238 horizontal edge and vice versa .
2 . let @xmath240 such that @xmath241 so that the @xmath231-entry of the block matrix @xmath242 is zero and @xmath243 + put @xmath244 and @xmath245 where @xmath246 is the sub diagram of @xmath247 and every @xmath238 through class of @xmath248 is replaced by a @xmath238 horizontal edge and vice versa .
3 . let @xmath249 such that @xmath250 so that the @xmath231-entry of the block matrix @xmath251 is zero and @xmath252 + put @xmath253 and @xmath254 where @xmath255 is the sub diagram of @xmath256 and every through class of @xmath257 is replaced by a horizontal edge and vice versa .
[ p2.16]*(@xcite , proposition 3.27 ) * let @xmath258 and @xmath259 be as in notation [ n2.15 ] . 1 .
after performing the column operations to eliminate the non - zero entries which lie above corresponding to the diagrams coarser than @xmath200 , 2 . after performing the column operations to eliminate the non - zero entries which lie above corresponding to the diagrams coarser than @xmath215 , + then the @xmath231-entry of the block matrix @xmath260 for @xmath261 and the block matrix @xmath262 for @xmath263
is replaced by 1 .
@xmath264 \underset{m = t_2}{\overset{r_2 - 1}{\prod } } [ x-(s_2+m)]$ ] if @xmath265 and @xmath266 2 .
@xmath267 $ ] if @xmath268 and @xmath269 3 . @xmath270 $ ] if @xmath152 and @xmath271 3 . after performing the column operations to eliminate the non - zero entries which lie above corresponding to the diagrams coarser than @xmath272 ,
then the @xmath231-entry is replaced by + @xmath273.$ ] [ t2.17]*(@xcite , theorem 3.29 ) * 1 .
let @xmath274 be the matrix similar to the gram matrix @xmath188 of the algebra of @xmath3-relations which is obtained after the column operations and the corresponding row operations on @xmath275 then + @xmath276 2 .
let @xmath277 be the matrix similar to the gram matrix @xmath204 of signed partition algebras which is obtained after the column operations and the corresponding row operations on @xmath278 then + @xmath279 + where 1 .
the diagonal element of @xmath280 is given by + @xmath281 \underset{m=0}{\overset{r_2 - 1}{\prod } } [ x-(s_2+m ) ] & if & r_1 \geq 1 , r_2 \geq 1 \\ 2 . & \underset{j=0}{\overset{r_1 - 1}{\prod } } [ x^2-x-2(s_1+j ) ] & if & r_2 = 0 \\ 3 . & \underset{m=0}{\overset{r_2 - 1}{\prod } } [ x-(s_2+m ) ] & if & r_1 = 0 \end{array } $ ] 2 .
the entry @xmath282 of the block matrix @xmath283 is replaced by + @xmath284 \underset{m=0}{\overset{r_2-t_2 - 1}{\prod } } [ x-(s_2 + t_2 + m ) ] & & \\ & \hspace{9cm}\text { if } r_1 \geq 1 , r_2 \geq 1 & & \\ 2 . &
( -1)^{t_1 } \ 2^{t_1 } \ ( t_1 ) ! \ \underset{j=0}{\overset{r_1-t_1 - 1}{\prod } } [ x^2-x-2(s_1 + t_1 + j ) ] \ \ \ \ \ \ \ \ \text { if } \ \
r_2 = 0 \\ 3 . & ( -1)^{t_2 } \ ( t_2 ) !
\underset{m=0}{\overset{r_2-t_2 - 1}{\prod } } [ x-(s_2 + t_2 + m ) ] \ \ \ \ \hspace{2 cm } \
\text { if } \ \
r_1 = 0 \end{array } $ ] + whenever @xmath199 and @xmath200 can be defined as in notation [ n2.15](a ) , ( b ) and proposition [ p2.16](a ) , ( b ) .
all other entries of the block matrix @xmath280 are zero .
+ the underlying partitions of the diagrams corresponding to the entries of the block matrix @xmath285 are @xmath286 ^ 1 [ 2 \lambda_2]^2 [ \lambda_3 ^ 2 ] [ 2 \lambda_4]^4 $ ] with @xmath287 such that atleast one of @xmath288 is greater than @xmath41 for @xmath289 and @xmath290 3 .
let @xmath291 where the partition @xmath292 ^ 1 , [ 1^{s_2}]^2 , [ 1^{r_1}]^3 , [ 1^{r_2}]^4)$ ] such that @xmath293 and @xmath291 is the block sub matrix corresponding to the diagrams whose underlying partition is @xmath294 1 .
the @xmath295-entry @xmath296 of the matrix @xmath291 is replaced by + @xmath297 \underset{m=0}{\overset{r'_2 - 1}{\prod } } [ x-(s_2+m ) ] + \underset{m=0}{\overset{k - s_1-s_2 - 1}{\prod } } [ x-(s_2+m)]$ ] + where @xmath298 and @xmath299 2 .
the zero in the @xmath231-entry is replaced by + @xmath300 \underset{m=0}{\overset{r'_2-t_2 - 1}{\prod } } [ x-(s_2+m+t_2 ) ] + \underset{m=0}{\overset{k - s_1-s_2 - 1}{\prod } } [ x-(s_2+m)]$ ] + where @xmath214 and @xmath215 are as in notation [ n2.15](b ) and proposition [ p2.16](b ) where @xmath301 and @xmath302 3 .
if @xmath303 then the @xmath231-entry is replaced by + @xmath304 $ ] + where @xmath305 and @xmath306 where @xmath301 and @xmath302 + all other entries of the block matrix @xmath291 are zero .
let @xmath307 be the matrix similar to the gram matrix @xmath308 which is obtained after the column operations and the row operations on @xmath309 then + @xmath310 + where 1 .
the diagonal element of @xmath311 is given by + @xmath312 $ ] 2 .
the entry @xmath282 of the block matrix @xmath311 is replaced by + @xmath313 $ ] + whenever @xmath225 and @xmath272 can be defined as in of notation [ n2.15](c ) and proposition [ p2.16](c ) .
all other entries of the block matrix @xmath311 are zero .
in this section , we introduce the symmetric diagram matrices @xmath0 of size @xmath1 based on the diagrams . by induction , we compute the eigenvalues of the symmetric diagram matrices @xmath0 using elementary row and column operations .
[ d3.2 ] fix @xmath126 and @xmath314 choose a @xmath315 such that @xmath74 has @xmath316-connected components where @xmath317 denote the set of all equivalence relations on @xmath318 we shall draw a diagram @xmath18 graphically using the graph @xmath74 with @xmath126 through classes as follows : 1 .
draw @xmath74 in the top row and a copy of @xmath74 denoted by @xmath75 in the bottom row .
2 . among the @xmath316 connected components in the top row ,
choose @xmath126 connected components and join each connected component with the respective connected component in the bottom row by vertical edges .
the resultant diagram is denoted by @xmath25 [ d3.3 ] let @xmath319 denote the collection of all diagrams with @xmath126 through classes mentioned as above in definition [ d3.2 ] and the number of such diagrams are denoted by @xmath320 [ e3.4 ] let @xmath321 and @xmath322 the diagrams with three through classes corresponding to the set partition @xmath323 are [ d3.5 ] fix @xmath126 and @xmath314 consider the following variables @xmath324 define a matrix @xmath0 of size @xmath1 with the entries @xmath325 as follows : @xmath326 with @xmath327 where @xmath328 denotes the number of horizontal edges in @xmath329 which are replaced by through classes in @xmath330 and vice versa , @xmath331 [ r3.6 ] for the sake of convenience , we shall replace every through class in the above mentioned diagram by a through class @xmath332 obtained by joining a single vertex in the top row and the corresponding vertex in the bottom row .
similarly , every horizontal edge in the top and bottom row is replaced by a vertex @xmath333 respectively .
[ l3.7 ] the matrix @xmath0 is symmetric .
since the top and bottom row of the diagrams in @xmath319 are the same , the matrix @xmath0 is symmetric .
[ d3.8 ] the matrix @xmath0 is called as * symmetric diagram matrix*. the number of times @xmath334 occurs in any row or column of a symmetric diagram matrix @xmath0 of size @xmath1 is given by @xmath335 the eigenvalues of the symmetric diagram matrix @xmath0 can be obtained using induction .
the main theorem of this paper gives all the distinct eigenvalues of the symmetric diagram matrix @xmath336 [ t3.9 ] the set of all distinct eigenvalues of the symmetric diagram matrix @xmath0 of size @xmath1 with entries @xmath337 are given by @xmath338 x_{\text{min}\{s , r\}-t}\ ] ] for all @xmath339 [ l3.10 ] if @xmath340 are @xmath341-variables and if @xmath342 then @xmath343 consider @xmath344 \ a_{t - j } + ( -1)^{l + 1 } \
a_{t-(l+1 ) } \\ & = & \underset{j=0}{\overset{l+1}{\sum } } ( -1)^j \ { _ { ( l+1)}}c_j \ a_{t - j } \\ & = & a_t^{l+1}\end{aligned}\ ] ] [ c3.11 ] in particular , if @xmath345 for @xmath346 then @xmath347 the proof follows from induction and lemma [ l3.10 ] . [ d3.12 ]
let @xmath348 with @xmath349 define , @xmath350 for all @xmath351 and the @xmath352 connected component of @xmath18 consists of two dots for all @xmath353 in particular , if @xmath354 and @xmath355 then @xmath356 the first connected component of @xmath18 is a through class @xmath357 and the @xmath358 connected component consists of two dots @xmath359 [ l3.13 ] let @xmath360 be as in definition [ d3.12 ] .
then @xmath361 in particular , @xmath362 the proof follows from the definition of @xmath363 we shall arrange the diagrams in @xmath319 as follows : 1 .
the collection of diagrams whose first connected component is a through class and the last connected component is a horizontal edge is denoted by @xmath364 .
also , @xmath365 and the diagrams in @xmath366 look like + + the diagrams in @xmath364 are indexed inductively as follows : + suppose @xmath367 is the @xmath368 diagram in @xmath369 , and which is the the subdiagram of @xmath329 lying inbetween the first and last connected component will be the @xmath368 diagram in @xmath364 inductively .
2 . for @xmath370 let @xmath371 then @xmath372 is the diagram same as @xmath329 except they differ at the first and last connected component of @xmath329 where the first connected component of @xmath372 is a horizontal edge and the last connected component is a through class . collection of such diagrams is denoted by @xmath373 the diagram @xmath372 will look like + 3 .
for @xmath374 the collection of diagrams whose first and last connected components are through classes is denoted by @xmath375 the diagrams in @xmath376 will look like + 4 . for @xmath377 the collection of all diagrams
whose first and last connected components are horizontal edges is denoted by @xmath378 the diagrams in @xmath379 will look like + [ l3.14 ] let @xmath380 , @xmath381 and @xmath382 then 1 .
@xmath383 2 .
@xmath384 * proof of ( i ) : * let @xmath380 and @xmath385 where @xmath386 denotes the number of horizontal edges in @xmath329 which are replaced by through classes in @xmath330 and vice versa .
from the definition of @xmath387 we know that the first and last connected components of @xmath329 and @xmath330 are the same . by the definition of @xmath388
we know that for every @xmath371 there exists a unique @xmath389 which differs only in the first and last connected component . also , the first and last connected components of @xmath372 and @xmath390 are the same .
thus , @xmath383 * proof of ( ii ) : * by the definition of @xmath387 the @xmath386 number of horizontal edges which are replaced by the through classes and vice versa lies in between the first and last connected components of @xmath329 and @xmath330 . also , the diagram @xmath389 is same as the diagram @xmath391 except the first and last connected component .
therefore , including the first and last connected component there will be @xmath392 number of horizontal edges in @xmath372 which are replaced by through classes in @xmath330 and vice versa .
thus , @xmath393 [ l3.15 ] let @xmath391 and @xmath394 then @xmath395 let @xmath396 where @xmath386 denotes the number horizontal edges in @xmath329 which is replaced by through classes and vice versa . from the definition of @xmath397 , we know that the diagram @xmath372 is same as the diagram @xmath391 except at the first and last connected components where the first(last ) connected component of @xmath329 is a through class(horizontal edge ) but the first(last ) connected component of @xmath372 is a horizontal edge(through class )
. @xmath398 denotes the entry corresponding to the product of diagrams @xmath372 and @xmath330 then by the definition of @xmath397 and @xmath399 we have , @xmath395 we shall now prove the main theorem of this paper using the arrangement of diagrams given above and induction .
* proof of main theorem : * the proof of the main theorem is using induction on the size of the matrix . * case ( i ) : * let @xmath400 and by definition [ d3.5 ] , the entries of the symmetric diagram matrix @xmath401 are @xmath402 where @xmath403
apply the following row operation and column operation on @xmath404 @xmath405 where the diagrams @xmath18 and @xmath406 differ only at the first and last connected component .
the first connected component of @xmath407 is through class(horizontal edge ) and the last connected component of @xmath407 is horizontal edge(through class ) .
the reduced matrix is @xmath408 thus , the eigenvalues of the symmetric diagram matrix @xmath401 are @xmath409 and @xmath410 * case ( ii ) : * let @xmath411 and by definition [ d3.5 ] , the entries of the symmetric diagram matrix @xmath412 are @xmath413 where @xmath414
apply the following row and column operation on @xmath415 @xmath405 for all @xmath416 and @xmath417 which differs only at the first and last connected component as in definition [ d3.12 ] .
then the reduced matrix is as follows : @xmath418 where @xmath419 is a symmetric diagram matrix of size @xmath420 with @xmath421 and @xmath422 and @xmath423 the diagrams corresponding to the entries of the symmetric diagram matrix @xmath424 belong to @xmath425 and the diagrams corresponding to the entries of the matrix @xmath426 belong to @xmath427 and @xmath428 using induction , the eigenvalues of the symmetric diagram matrix @xmath429 are @xmath430 and @xmath431 i.e. , @xmath432 and @xmath433 apply the following row and column operation on the matrix @xmath434 @xmath435 for the diagrams @xmath436 and @xmath437 thus , the reduced matrix is as follows : @xmath438 where @xmath439 is a matrix of size @xmath41 and @xmath440 the diagram corresponding to the entry of the matrix @xmath441 belong to @xmath442 and the diagrams corresponding to the entries of the matrix @xmath443 belong to @xmath444 and @xmath445 the eigenvalue of the matrix @xmath446 is @xmath433 apply the following row and column operations on the matrix @xmath447 @xmath448 thus , the reduced matrix is as follows : @xmath449 where @xmath450 and @xmath451 which is same as @xmath452 therefore , the eigenvalues of the symmetric diagram matrix @xmath0 are @xmath453 and @xmath454 in general , we shall do this for the diagrams in @xmath319 and shall find the eigenvalues of the symmetric diagram matrix @xmath0 using induction .
* step i : * we shall split @xmath319 into four subsets as follows : @xmath455 such that for all @xmath456 there exist a unique @xmath457 which differs only at the first and @xmath358 connected components where the first connected component of @xmath407 is through class ( horizontal edge ) and the @xmath358 connected component of @xmath407 is horizontal edge ( through class ) . clearly , @xmath458 and @xmath459
apply the following row and column operations on the symmetric diagram matrix @xmath0 of size @xmath460 @xmath461 and @xmath462 using lemmas [ l3.14 ] and [ l3.15 ] , the reduced matrix is as follows : @xmath463 where @xmath464 is a symmetric diagram matrix of size @xmath465 with entries @xmath466 and @xmath467 the diagrams corresponding to the entries of the symmetric diagram matrix @xmath424 belong to @xmath468 using induction the eigenvalues of the symmetric diagram matrix @xmath464 of size @xmath465 are given as @xmath469}}c_{(t - j ) } \ { _ { [ ( r-1)-(l-1)]}}c_{(t - j)}\right\ } y_{\text{min } \{s-1 , r-1\}-t}\ ] ] for all @xmath470 substitute @xmath471 in expression ( [ e3.2 ] ) we get , @xmath472}}c_{(t - j ) } \ { _ { [ ( r-1)-(l-1)]}}c_{(t - j)}\right\ } \\ & & \hspace{8 cm } \left\ { x_{\text{min}\{s-1,r-1\}-t+1 } - x_{\text{min}\{s-1,r-1\}-t } \right\ } \\ & = & \underset{t=0}{\overset{\text{min}\{s-1,r-1\}}{\sum } } \left\ { \underset{j=0}{\overset{l-1}{\sum } } ( -1)^j \ { _ { ( l-1)}}c_j \ { _ { ( s - l)}}c_{(t - j ) } \ { _ { ( r - l)}}c_{(t - j ) } - \underset{j=0}{\overset{l-1}{\sum } } ( -1)^j \ { _ { ( l-1)}}c_j \ { _ { ( s - l)}}c_{(t-1-j ) } \ { _ { ( r - l)}}c_{(t-1-j)}\right\}\\ & & \hspace{13 cm } x_{\text{min}\{s-1 , r-1\}-t+1}\\ & & \hspace{6cm}(\text{by collecting the coefficients } x_{\text{min}\{s-1 , r-1\}-t+1})\\ & = & \underset{t=0}{\overset{\text{min}\{s-1,r-1\}}{\sum } } \left\ { \underset{j=0}{\overset{l-1}{\sum } } ( -1)^j \ { _ { ( l-1)}}c_j \ { _ { ( s - l)}}c_{(t - j ) } \ { _ { ( r - l)}}c_{(t - j ) } - \underset{j'=1}{\overset{l}{\sum } } ( -1)^{j'-1 } \ { _ { ( l-1)}}c_{(j'-1 ) } \ { _ { ( s - l)}}c_{(t - j ' ) } \ { _ { ( r - l)}}c_{(t - j')}\right\}\\ & & \hspace{13 cm } x_{\text{min}\{s-1 , r-1\}-t+1}\\ & = & \underset{t=0}{\overset{\text{min}\{s-1,r-1\}}{\sum } } \left\ { { _ { ( s - l)}}c_t \ { _ { ( r - l)}}c_t + \underset{j=0}{\overset{l-1}{\sum } } ( -1)^j \ { _ { ( l-1)}}c_j \ { _ { ( s - l)}}c_{(t - j ) } \ { _ { ( r - l)}}c_{(t - j ) } + ( -1)^l \ { _ { ( s - l)}}c_{(t - l ) } \ { _ { ( r - l)}}c_{(t - l)}\right\ } \\ & & \hspace{13 cm } x_{\text{min}\{s-1 , r-1\}-t+1 } \end{aligned}\ ] ] @xmath473 thus , the eigenvalues of the symmetric diagram matrix @xmath424 of size @xmath474 are given by @xmath475 therefore , the eigenvalues of the symmetric diagram matrix @xmath476 of size @xmath465 are @xmath477 the number of distinct eigenvalues of the symmetric diagram matrix @xmath0 so far computed which is given in ( [ e3.4 ] ) are min @xmath478 we shall now prove that the eigenvalues of the submatrix @xmath426 are also as given in expression ( [ e ] ) .
* step 2 : * the diagrams corresponding to the entries of the matrix @xmath479 belong to @xmath480 for @xmath481 where @xmath482 and @xmath483 we shall apply the following row and column operations on @xmath484 @xmath485 and @xmath486 for all @xmath487 using lemmas [ l3.14 ] and [ l3.15 ] and applying suitable row and column operations to the reduced matrix looks like , @xmath488 the diagrams corresponding to the entries of the matrix @xmath446 belong to @xmath489 for all @xmath487 the size of the matrix @xmath446 is @xmath490 now , we shall show that the eigenvalues of the matrix @xmath441 belong to the collection of all eigenvalues of the matrix @xmath424 obtained in step 1 . we know
that the diagrams corresponding to the entries of the symmetric diagram matrix @xmath424 obtained in step @xmath41 belong to @xmath468 apply the following row and column operations on the symmetric diagram matrix @xmath491 @xmath492 for all @xmath493 and @xmath494 using induction on the number of through classes the reduced matrix is as follows : @xmath495 where the diagrams corresponding to the entries of the matrix @xmath496 belong to @xmath497 and the diagrams corresponding to the entries of the matrix @xmath498 belong to @xmath499 and @xmath500 the size of the matrix @xmath501 is @xmath474 and the size of the matrix @xmath498 is @xmath502 since , @xmath501 and @xmath503 are the submatrices of the symmetric diagram matrix @xmath464 , by induction the eigenvalues of the matrices @xmath501 and @xmath498 belong to the collection of all eigenvalues given in expression [ e ] .
it is clear from the definition of @xmath360 defined in definition [ d3.12 ] that the submatrices @xmath498 and @xmath441 are the same .
hence the eigenvalues of the matrix @xmath498 and @xmath441 are the same .
the diagrams corresponding to the entries of the matrix @xmath443 belong to @xmath504 and @xmath505 for @xmath487 * step j : * in general , the diagrams corresponding to the entries of the matrix @xmath506 obtained in step @xmath507 belong to @xmath508 and @xmath509 with @xmath510 for all @xmath511 apply the following row and column operations on @xmath512 @xmath492 for all @xmath513 and @xmath514 using lemmas [ l3.14 ] and [ l3.15 ] and interchanging the rows and columns suitably the reduced matrix is as follows : @xmath515 the diagrams corresponding to the entries of the matrix @xmath516 belong to @xmath517 for all @xmath518 the size of the matrix @xmath516 is @xmath519 now , we shall show that the eigenvalues of the matrix @xmath520 and the eigenvalues of the matrix @xmath521 obtained from @xmath503 inductively as in step 2 are the same . using induction
, we know that the diagrams corresponding to the entries of the symmetric diagram matrix @xmath522 obtained in step @xmath523 belong to @xmath524 and @xmath525 for all @xmath526 apply the following row and column operations on the symmetric diagram matrix @xmath527 @xmath492 for all @xmath528 and @xmath529 using induction on the number of through classes , lemma [ l3.14 ] , [ l3.15 ] and interchanging rows and column suitably the reduced matrix is as follows : @xmath530 where the diagrams corresponding to the entries of the matrix @xmath531 belong to @xmath532 and the diagrams corresponding to the entries of the matrix @xmath521 belong to @xmath533 and @xmath534 for all @xmath511 the size of the matrix @xmath521 is @xmath535 since , @xmath536 and @xmath537 are the submatrices of the symmetric diagram matrix @xmath464 , the eigenvalues of the matrices @xmath536 and @xmath521 belong to the collection of all eigenvalues given in expression [ e ] .
it is clear from the definition of @xmath360 defined in definition [ d3.12 ] that the submatrices @xmath521 and @xmath538 are the same .
hence , eigenvalues of the matrix @xmath521 and @xmath520 are the same .
the diagrams corresponding to the entries of the matrix @xmath538 belong to @xmath539 and @xmath540 for @xmath518 some of @xmath541 and @xmath542 may be empty .
when the set is empty we leave that set and continue the process . also , we always consider the matrix with coefficient @xmath41 for the highest order in the determinant .
this process is continued till step @xmath543 if @xmath544 and @xmath545 if @xmath546 if @xmath547 then we apply the following row and column operations on @xmath548 @xmath549 and @xmath550 for all @xmath551 and @xmath552 and @xmath553 in both the cases ( i.e. , @xmath544 and @xmath547 ) , using lemmas [ l3.14 ] , [ l3.15 ] and by interchanging rows and columns suitably the reduced matrix is as follows : @xmath554 the diagrams corresponding to the entries of the matrix @xmath555 belong to @xmath480 for all @xmath556 finally , we shall apply the following row and column operations on the matrix @xmath557 1 . fix a @xmath558 + @xmath559 + where @xmath560 + @xmath561 for all @xmath560 2 .
fix a @xmath562 + @xmath559 + where @xmath563 + @xmath561 for all @xmath564 3 .
fix a @xmath565 + @xmath559 + where @xmath566 + @xmath561 for all @xmath566 using induction , lemma [ l3.14 ] , lemma [ l3.15 ] and after applying suitable row and column operations on the reduced matrix , it becomes , @xmath567 where the size of the matrix @xmath568 is @xmath569 and the matrix @xmath570 is one among the submatrices obtained after applying row and column operations on the symmetric diagram matrix @xmath424 inductively .
thus , the eigenvalues of the matrix @xmath570 belong to the collection of all eigenvalues given in expression [ e ] .
now , we are left out to find the eigenvalues of the matrix @xmath571 for that we add all the entries to one column and subtract the corresponding row with other rows .
the reduced matrix is as follows : @xmath572 where @xmath573 is a @xmath574 matrix and @xmath575 is a @xmath576 matrix . since
, we have only performed addition on the columns the entry of the matrix @xmath573 is the sum of the entries of the symmetric diagram matrix @xmath0 of size @xmath577 which is given by @xmath578 we perform the following row and column operations on the matrix @xmath579 @xmath580 and @xmath581 thus , the eigenvalues of the matrix @xmath575 also belong to the collection of all eigenvalues given in expression [ e ] .
thus , we have computed all the @xmath582 number of distinct eigenvalues of the symmetric diagram matrix @xmath336 thus , the eigenvalues of the symmetric diagram matrix @xmath0 of size @xmath577 is given by @xmath583 for all @xmath584
* + * in this section , we compute the eigenvalues of gram matrices of partition algebras where the block submatrices of the gram matrix are realized as the direct product of the symmetric diagram matrices .
[ t4.1 ] 1 .
the set of all distinct eigenvalues of gram matrix @xmath219 of the partition algebra with entries @xmath585 are given by + @xmath586 x_{\text{min}\{s , r\}-t}$ ] + for all @xmath587 and @xmath588 where @xmath589 , k$ ] and @xmath126 are fixed integers .
2 . the eigenvalues of the gram matrix @xmath219 are integers which is given by + @xmath590 x_{\text{min}\{s , r\}-t } = \underset{i=0}{\overset{l-1}{\prod } } [ x-(s-1+i ) ] \underset{j=0}{\overset{\text{min}\{s , r\}-l-1}{\prod}}[x-(2s+j)]$ ] + @xmath584 * * proof of ( i):**let @xmath591 such that @xmath592 where @xmath593 and @xmath594 are as in notation [ n2.12 ] and definition [ d2.11 ] respectively .
we shall draw a diagram using the diagram @xmath595 with @xmath126 through classes as follows : 1 .
draw @xmath595 denoted by @xmath596 in the top row and a copy of @xmath595 denoted by @xmath597 in the bottom row .
2 . among the @xmath316 connected components in the top row
choose @xmath126 connected components and join each connected component in the top row with the respective connected component in the bottom row by vertical edges .
3 . denote the collection of such diagrams as @xmath598 in particular @xmath599 it is clear that the number of such diagrams with @xmath126 through classes is @xmath1 and @xmath600 where @xmath601 is as in notation [ n2.12 ] .
suppose @xmath602 then the entry in the gram matrix corresponding to the product @xmath603 is either @xmath604 with @xmath605 or @xmath606 mod @xmath294 * case ( i ) : * if the entry corresponding to the product @xmath603 in the gram matrix @xmath219 is @xmath607 with @xmath605 and while applying the column operations to eliminate the entries corresponding to the diagrams coarser than @xmath608 the entry @xmath607 becomes zero by lemma 3.20 in @xcite .
* case ( ii ) : * if the entry corresponding the product @xmath603 in the gram matrix @xmath219 is @xmath606 mod @xmath609 and it remains zero even after applying column operations to eliminate the entries corresponding to the diagrams coarser than @xmath608 by lemma 3.19 and lemma 3.23 in @xcite . rearranging the diagrams @xmath591 in such a way that @xmath610 where @xmath611 is a symmetric diagram matrix of size @xmath577 with entries @xmath585 and the diagrams corresponding to the entries of the symmetric diagram matrix @xmath0 belong to @xmath612 also ,
the by theorem [ t2.17](c ) @xmath613.$ ] by theorem [ t3.9 ] , the eigenvalues of the symmetric diagram matrix @xmath611 of size @xmath577 are given by @xmath586 x_{\text{min}\{s , r\}-t}$ ] for all @xmath584 thus , the eigenvalues of the block submatrix @xmath311 which is a direct product of symmetric diagram matrices @xmath611 are given by @xmath586 x_{\text{min}\{s , r\}-t}$ ] for all @xmath584 therefore , the eigenvalues of the gram matrix @xmath219 are given by @xmath586 x_{\text{min}\{s , r\}-t}$ ] for all @xmath614 and @xmath615 * proof of ( ii ) : * the proof of ( ii ) follows by comparing the coefficients on both the sides .
* + * [ t4.2 ] 1 .
the set of all distinct eigenvalues of gram matrices @xmath188 of the algebra of @xmath3-relations with entries @xmath616 and @xmath617 are given by 1 .
@xmath618 x_{\text{min}\{s_1 , r_1\}-t}$ ] + for all @xmath619 and @xmath620 + where @xmath621 , k , s_1 $ ] and @xmath96 are fixed integers 2 .
@xmath622 x'_{\text{min}\{s_2 , r_2\}-t}$ ] + for all @xmath623 and @xmath624 + where @xmath625 , k , s_1 $ ] and @xmath96 are fixed integers .
the set of all distinct eigenvalues of block submatrices @xmath626 of the gram matrix @xmath204 of signed partition algebra with entries @xmath616 and @xmath617 are given by 1 .
@xmath618 x_{\text{min}\{s_1 , r_1\}-t}$ ] + for all @xmath619 and @xmath620 + where @xmath621 , k , s_1 $ ] and @xmath96 are fixed integers 2 .
@xmath622 x'_{\text{min}\{s_2 , r_2\}-t}$ ] + for all @xmath623 and @xmath627 + where @xmath625 , k , s_1 $ ] and @xmath96 are fixed integers .
since the @xmath87-connected components can not be replaced by @xmath3-connected components , we can rearrange the diagrams such that the block matrices @xmath628 becomes the tensor product of matrices say @xmath629 respectively .
let @xmath630 such that @xmath631 where @xmath632 is as in definition [ d2.11 ] , @xmath633 and @xmath634 are as in notation [ n2.12 ] .
we shall draw a diagram using the diagram @xmath635 with @xmath636 through classes respectively as follows . 1 .
draw @xmath637 denoted by @xmath74 in the top row and a copy of @xmath638 denoted by @xmath639 in the bottom row .
2 . fix the positions of @xmath3-connected component and among the @xmath640 number of pairs @xmath87-connected components in the top row choose @xmath95 pairs of @xmath87 connected components and join each pair of @xmath87-connected component in the top row with the respective pair of @xmath87-connected component in the bottom row by vertical edges .
+ similarly , fix the positions of @xmath87-connected components and among the @xmath641 number of pairs @xmath3-connected components in the top row choose @xmath96 number of @xmath3-connected components and join each @xmath3-connected component in the top row with the respective @xmath3-connected component in the bottom row by vertical edges .
denote the collection of diagrams obtained by fixing @xmath3-connected components as @xmath642 and denote the collection of diagrams obtained by fixing @xmath87-connected components as @xmath643 it is clear that the number of such diagrams with @xmath636 through classes is @xmath644 and @xmath645 where @xmath633 and @xmath634 are as in notation [ n2.12 ] .
suppose @xmath646 then the entry corresponding to the product @xmath647 in the gram matrix is either @xmath648 with @xmath649 and @xmath650 or @xmath651 mod @xmath652 * case ( i ) : * if the entry corresponding to the product @xmath647 is @xmath648 with @xmath649 and @xmath650 then it becomes zero after applying column operations to eliminate the entries corresponding to the diagrams coarser than @xmath653 by lemma 3.20 in @xcite .
* case ( ii ) : * if the entry corresponding to the product @xmath647 is @xmath651 mod @xmath37 then it remains zero even after applying the column operations to eliminate the entries corresponding to the diagrams coarser than @xmath653 by lemma 3.19 and lemma 3.23 in @xcite .
rearranging the diagrams @xmath654 in such a way that @xmath655 and @xmath656 where @xmath657 and @xmath658 are symmetric diagram matrices of size @xmath659 and @xmath660 respectively .
the entries of the symmetric diagram matrices @xmath657 and @xmath658 are @xmath616 and @xmath661 respectively and the diagrams corresponding to the entries of the symmetric diagram matrices @xmath657 and @xmath658 belong to @xmath662 and @xmath663 respectively .
similarly , we can rearrange the diagrams in signed partition algebra in such a way that @xmath664 and @xmath665 where @xmath666 and @xmath667 are symmetric diagram matrices of size @xmath659 and @xmath660 respectively .
the entries of the symmetric diagram matrices @xmath666 and @xmath667 are @xmath616 and @xmath661 respectively and the diagrams corresponding to the entries of the symmetric diagram matrices @xmath666 and @xmath667 belong to @xmath668 and @xmath669 respectively .
also , the by theorem [ t2.17](a ) @xmath670 $ ] and @xmath671.$ ] by theorem [ t3.9 ] , the eigenvalues of the symmetric diagram matrix @xmath672 of size @xmath659 are given by @xmath618 x_{\text{min}\{s_1 , r_1\}-t}$ ] for all @xmath673 by theorem [ t3.9 ] , the eigenvalues of the symmetric diagram matrix @xmath674 of size @xmath660 are given by @xmath622 x'_{\text{min}\{s_2 , r_2\}-t}$ ] for all @xmath675 thus , the eigenvalues of the block submatrix @xmath676 which is a direct product of symmetric diagram matrices @xmath677 are given by @xmath618 x_{\text{min}\{s_1 , r_1\}-t}$ ] for all @xmath673 thus , the eigenvalues of the block submatrix @xmath678 which is a direct product of symmetric diagram matrices @xmath679 are given by @xmath680 x'_{\text{min}\{s_2 , r_2\}-t}$ ] for all @xmath675 therefore , the eigenvalues of the gram matrix @xmath188 of the algebra of @xmath3-relations are given by 1 .
@xmath618 x_{\text{min}\{s_1 , r_1\}-t}$ ] + for all @xmath619 and @xmath620 + where @xmath621 , k , s_1 $ ] and @xmath96 are fixed integers 2 .
@xmath622 x'_{\text{min}\{s_2 , r_2\}-t}$ ] + for all @xmath623 and @xmath624 + where @xmath625 , k , s_1 $ ] and @xmath96 are fixed integers .
to compute the eigenvalues of the symmetric diagram matrix @xmath681 of size @xmath682 , we need the eigenvalues of the symmetric diagram matrices @xmath683 and @xmath684 of sizes @xmath569 and @xmath685 respectively .
first , we shall compute the eigenvalues of the symmetric diagram matrix @xmath683 of size @xmath686 where @xmath698 is a symmetric diagram matrix of size @xmath569 with @xmath699 and @xmath700 using induction the eigenvalues of the symmetric diagram matrix @xmath424 are @xmath701 and @xmath702 i.e. , @xmath703 and @xmath704 respectively . where @xmath714 , @xmath715 is same as @xmath441 which is a submatrix of the symmetric diagram matrix @xmath424 obtained in step 1
. therefore , the eigenvalues of @xmath716 are same as the eigenvalues of @xmath452 where @xmath573 is a @xmath574 matrix whose entry is sum of the entries of the symmetric diagram matrix @xmath684 i.e. , @xmath720 and @xmath721 which is again a submatrix of the symmetric diagram matrix @xmath424 obtained in step 1 .
therefore , the eigenvalues of @xmath575 are same as the eigenvalues of @xmath452 where @xmath740 is same as @xmath498 a submatrix of @xmath424 obtained after applying the row and column operations.thus , the eigenvalues of the matrix @xmath441 are same as the eigenvalues of matrix @xmath452 1 .
fix a @xmath761 + @xmath559 + where @xmath762 + @xmath561 for all @xmath762 2 .
fix a @xmath763 + @xmath559 + where @xmath764 + @xmath561 for all @xmath765 3 . fix a @xmath766 + @xmath559 + where @xmath767 + @xmath561 for all @xmath767 where @xmath573 is a @xmath574 matrix whose entry is sum of the entries of the symmetric diagram matrix @xmath681 i.e. , @xmath773 and @xmath774 we perform the following row and column operation on @xmath579 j. j. graham and g. i. lehrer , _ cellular algebras _ , inventiones mathematicae , 123 , 1 - 34(1996 ) .
r. m. green , tabular algebras and their asymptotic versions , j. algebra * 252 * ( 2002 ) , no . 1 , 2764 .
r. m. green and p. p. martin , constructing cell data for diagram algebras , j. pure appl .
algebra * 209 * ( 2007 ) , no . 2 , 551569 .
arun ram , tom halverson , _ the partition algebras _ , european journal of electronics .
v. f. r. jones , the potts model and the symmetric group , in _ subfactors ( kyuzeso , 1993 ) _
, 259267 , world sci .
publ . , river edge , nj .
g. james and a. kerber , _ the representation theory of the symmetric group _ , encyclopedia of mathematics and its applications , addison- wesley , 16 ( 1981 ) .
kosuda , m.irreducible representation of the party algebra .
osaka.j.math.43 , 431 - 474(2006 ) . n. karimilla bi , m. parvathi _ gram matrices and stirling numbers of a larger class of diagram algebras_(communicated ) . n. karimilla bi , cellularity of a larger class of diagrams accepted in kyungpook mathematical journal .
p. martin and h. saleur , algebras in higher - dimensional statistical mechanics the exceptional partition ( mean field ) algebras , lett .
* 30 * ( 1994 ) , no . 3 , 179185 .
p. martin , the structure of the partition algebras , j. algebra * 183 * ( 1996 ) , no . 2 , 319358 .
p. martin , the partition algebra and the potts model transfer matrix spectrum in high dimension , j. phys .
a * 33 * ( 2000 ) 3669 3695 .
martin , potts models and related problems in statistical mechanics , series on advances in stastical mechanics , vol.5 , world sientific publishing co. inc .
, teaneck nj , 1991 .
r. dipper and g. james , representations of hecke algebras of general linear groups , proc .
london math .
( 3 ) * 52 * ( 1986 ) , no . 1 , 2052 . b. e. sagan , _ the symmetric group _ , second edition , graduate texts in mathematics , 203 , springer , new york , 2001 .
tanabe , k. _ on the centralizer algebra of the unitary reflection group @xmath780 .
_ nagoya math .
j. 148 , 113 - 126 ( 1997 ) . | in this paper , we introduce symmetric diagram matrices @xmath0 of size @xmath1 whose entries are @xmath2 .
we compute the eigenvalues of symmetric diagram matrices using elementary row and column operations inductively . as a byproduct
, we obtain the eigenvalues of gram matrices of a larger class of diagram algebras like the signed partition algebras , algebra of @xmath3 relations and partition algebras .
* keywords : * eigenvalues , symmetric diagram matrices , gram matrices , partition algebras , signed partition algebras and algebra of @xmath3-relations .
* mathematics subject classification(2010 ) . * 15b99 , 16z05 . |
flipped su(5 ) enthusiasts keep discovering hidden treasures , even after 10 years from its birth @xcite .
as is well known , the model attains its highest relevance in strings : efforts by several groups using different approaches have not ( yet ? ) yielded appealing string guts " [ so(10 ) ] . among level - one kac - moody models , only flipped su(5 ) unifies su(3 ) and su(2 ) , providing an explanation for the lep scale " [ @xmath4 .
the discrepancy between observed " and predicted unification scales
versus @xmath6 seems to have only way out : extra intermediate - scale states @xcite .
this solution was realized early on in stringy flipped su(5 ) @xcite .
here we summarize how this scenario may be achieved in practice @xcite , including the prediction for @xmath0 @xcite , and also discuss the latest flipped " goodie : a leptophobic @xmath1 @xcite .
: + @xmath7 ; @xmath8 ; @xmath9 @xmath10 + @xmath7 ; @xmath11 + : + @xmath12 ; @xmath13 @xmath14 ; @xmath15 + : @xmath16 the vevs @xmath17 break @xmath18 down to @xmath19 .
+ doe / er/40717 - 30 + hep - ph/9607231 : @xmath20 @xmath21 the triplets get heavy , while the doublets remain light ( missing partner mechanism " ) .
+ : @xmath22 : the gut couplings @xmath23
, @xmath24 entail @xmath25 see - saw mechanism : @xmath26 + good for msw mechanism , @xmath27 hot dark matter , and ( @xmath28 ) baryogenesis .
+ : mediated by @xmath29 gut gauge bosons , the mode @xmath30 may be observable at superkamiokande .
+ : very suppressed since no @xmath31 mixing exists , even though @xmath31 are heavy via doublet - triplet splitting , @xmath32
starting from the low - energy standard model gauge couplings , and evolving them from low to high energies , first @xmath33 and @xmath34 unify at @xmath35 @xcite @xmath36 the hypercharge does not unify at @xmath35 : @xmath37 above @xmath35 the gauge group is su(5)@xmath38u(1 ) .
stringy unification occurs at @xmath39 there is an @xmath40 @xmath41 solving for @xmath34 , to lowest order : @xmath42 compare with su(5 ) where @xmath43 @xcite @xmath44 what happens at next - to - leading order ?
@xmath45 decreasing @xmath46 increases @xmath0 [ avoid ! ] : @xmath47 ; @xmath48 ( light susy thresholds ) ; @xmath49 ( gut thresholds ) @xmath50\ ] ] since there is no problem with proton decay , @xmath49 can be negative .
we obtain @xmath0 as low as 0.108 ( see fig .
[ fig : sin2r ] ) .
however , decreasing @xmath35 decreases the proton lifetime @xmath51 the present lower bound @xmath52 implies @xmath53 ( see fig .
[ fig : taup ] ) . if @xmath54 then @xmath30 may be observable at superkamiokande ( which should have a sensitivity of @xmath55 ) .
this is in contrast with minimal su(5 ) , where the preferred mode is @xmath56 .
string construction in fermionic formulation @xcite + : @xmath57 + @xmath58=su(5)@xmath38u(1 ) ; @xmath59=so(10)@xmath38su(4 ) ; @xmath60=u(1)@xmath61 + + observable sector : @xmath62 & \bar f^{\{2,3,5\}}[\bar 5]&{\ell^c}^{\{2,3,5\}}[1]\\ \bar f^{\{4,5\}}[\overline{10}]\\ h^{\{1,2,3,45\}}[5]&\bar h^{\{1,2,3,45\}}[\bar5 ] \end{array}\ ] ] singlets : 20 charged under u(1 ) s , 4 neutral hidden sector : @xmath63\ of\ so(10)\\ \rm d^{\{1,2,3,4,5,6,7\}}&\rm [ 6]\ of\ su(4)\\ \rm \f^{\{1,2,3,4,5,6\ } } & \rm [ 4]\ of\ su(4)\\ \rm \fb^{\{1,2,3,4,5,6\ } } & \rm [ \bar4]\ of\ su(4 ) \end{array}\ ] ] the @xmath64 fields carry @xmath65 electric charges and exist only in hadron - like _
cryptons_. the cubic and non - renormalizable terms in the superpotential have been calculated @xcite , and more recently also the khler potential @xcite .
the properties of the khler potential illuminate the vacuum energy ( which vanishes at tree level and possibly also at one loop ) and determine the pattern of soft - supersymmetry - breaking masses , which has distinct experimental consequences @xcite .
assume that @xmath66 breaks as in standard flipped su(5 ) case .
cancellation of @xmath67 is consistent with @xmath68 correct @xmath46 and @xmath34 obtained because of extra , present in string massless spectrum .
string unification occurs at @xmath6 .
this requires @xmath69 , which can be generated via vevs of hidden matter fields .
@xcite + @xmath70 affect running of u(1 ) down to @xmath71 , where @xmath72 ; @xmath73 depends on string spectrum of @xmath74 , and on actual decoupling of particles between @xmath75 and @xmath73 [ tricky ] .
extra @xmath76 affect running of su(5)@xmath38u(1 ) down to @xmath77 . naively , if @xmath78 , a non - renormalizable term @xmath79 implies @xmath80 .
but in strings @xmath81 is very unlikely ; @xmath82 is more natural . in the actual string model we have a quintic term [ and @xmath82 ] : @xmath84 , where the cancellation of @xmath67 implies @xmath85 . with massless flavors ( @xmath82 )
one expects @xmath86 .
aharony , et .
al . studied @xmath87 with @xmath88 massless " flavors with supersymmetry - breaking scalar masses @xcite .
supersymmetry - breaking masses ( @xmath89 ) yield finite condensates @xmath90 ^{(n_c - n_f)/2(2n_c - n_f)}\ ] ] in our case ( @xmath91 , @xmath92 ) we obtain @xmath93 and we can @xmath77 from first principles @xmath94 this result allows self - consistent string unification .
the results for the various scales as a function of @xmath0 are shown in fig .
[ fig : scales ] . the full evolution of the gauge couplings from the weak scale to the string scale
is shown in fig .
[ fig : runnings ] for the preferred choices of @xmath95 and @xmath96 .
: original smoking gun " of string ; @xmath97 ` crisis ' has revived interest in @xmath1 models , although this time the @xmath1 must not couple to leptons .
leptophobia is _ natural _ in flipped su(5 ) @xcite @xmath98 if the leptons are uncharged , most quarks may be charged under @xmath99 .
compare with regular su(5 ) @xmath100 , where uncharged leptons imply uncharged quarks .
dynamic leptophobia ( via rge u(1 ) mixing ) is also possible , as in the @xmath101-model in @xmath102 @xcite .
any @xmath103-@xmath1 mixing shifts the usual @xmath103 couplings ( @xmath104 ) : @xmath105 , @xmath106 , where @xmath107 is the @xmath103-@xmath1 mixing angle ( small ) ; @xmath108 are the @xmath109 gauge couplings ; and @xmath110 the fermion couplings to the @xmath1 . in flipped su(5 ) we have [ @xmath111 , @xmath112 @xmath113 we can determine the first - order shifts in @xmath114 , @xmath115 , and @xmath116 , allowing for non - universal @xmath117 charges picked from @xmath118 this @xmath99 charge space satisfies specific requirements : the leptons ( in @xmath119 ) are uncharged ; one uncharged @xmath120 pair ( @xmath121 ) so that @xmath99 remains unbroken upon su(5)@xmath38u(1 ) breaking ; @xmath122 enforced ; extra @xmath120 to allow string unification .
the actual string model underlies these choices .
there are 13 possible charge assignments that can be made .
phenomenology demands @xmath123 , as the sm prediction and lep agree well . since @xmath124 and @xmath125 ( @xmath126 fixed to sm value )
, we demand @xmath127 . fig .
[ fig : rbghad ] shows @xmath128 versus @xmath129 .
an analogous plot for @xmath130 versus @xmath128 , demanding @xmath130 , @xmath128 shifts in opposite directions can be found in ref .
we should keep in mind that experimentally there appears to be a trend of @xmath126 converging to the standard model prediction and @xmath131 approaching it significantly .
consider @xmath132 , with @xmath133 , @xmath134 .
the anomalous combination is @xmath135 , with three orthogonal traceless combinations : @xmath136 ; @xmath137 ; @xmath138 .
the lepton charges under @xmath139 are @xmath140 ; @xmath141 .
+ there is a _ unique _
@xmath99 that is leptophobic @xmath142 and by construction @xmath143 .
higgs fields charged under @xmath99 exist ( @xmath103-@xmath1 mixing ) .
the d- and f - flatness conditions may be satisfied , leaving @xmath99 unbroken , but breaking the hidden group .
+ : @xmath144 should contain 3rd generation ( top yukawa ) ; @xmath121 neutral under @xmath99 : symmetry breaking ; @xmath145 : string unification ; @xmath97 inputs : four charge assignments allowed @xmath146 unlike any considered before .
? obtained from string !
top - quark yukawa coupling , and @xmath97 select scenario ( 11 ) uniquely @xmath147 : running of @xmath99 from @xmath148 up looks good : @xmath149 ( c.f .
@xmath150 ) .
sufficiently small @xmath103-@xmath1 mixing appears to require radiative @xmath99 symmetry breaking via singlet @xmath151 .
@xmath1 width and branching ratios for preferred case : @xmath152 @xmath153}\ ] ] experimental limits : @xmath154 average up / down ; multiply by @xmath155 , @xmath156 only limit from ua2 : @xmath157 , but only if @xmath158 .
@xmath1 contributes to top - quark cross section ( see fig . 3 in ref .
@xcite ) at a level that may be observable if @xmath159 .
parity - violating spin asymmetries at rhic may also show deviations from standard model expectations because of the @xmath160-channel exchange of our parity - violating @xmath1 . in sum , flipped su(5 )
continues to provide unsolicited solutions to unanticipated problems , as evidenced most recently by the self - consistent string unification and the possible existence of a leptophobic @xmath1 gauge boson .
99 for a recent review see and , hep - ph/9511266 .
k. dienes and a. faraggi , , .
j. l. lopez , d. v. nanopoulos , and k. yuan , nucl .
b * 399 * ( 1993 ) 654 .
j. l. lopez and d. v. nanopoulos , .
j. ellis , , and , .
j. l. lopez and d. v. nanopoulos , hep - ph/9605359 .
, , and k. yuan , ; and , hep - ph/9412332 . , , and a. zichichi , and .
aharony , et .
al . , . k. babu , c. kolda , and j. march - russell , hep - ph/9603212 . | we summarize recent developments in the prediction for @xmath0 , self - consistent string unification and the dynamical determination of mass scales , and leptophobic @xmath1 gauge bosons in the context of stringy flipped su(5 ) .
= 11 # 1#2 # 1#1| # 1| # 1 # 1#1 # 1#1 versim#1#2 # 1@xmath2 # 1@xmath3 # 1#2#3nucl .
phys .
b * # 1 * ( 19#2 ) # 3 # 1#2#3phys . lett .
b * # 1 * ( 19#2 ) # 3 # 1#2#3phys . rev .
d * # 1 * ( 19#2 ) # 3 # 1#2#3phys .
. lett . *
# 1 * ( 19#2 ) # 3 # 1#2#3phys . rep . *
# 1 * ( 19#2 ) # 3 # 1#2#3mod .
phys .
lett . a * # 1 * ( 19#2 )
# 3 # 1#2#3int .
j. mod .
phys . a * # 1 * ( 19#2 )
# 3 |
the growth of ice crystals and aggregate snowflakes in clouds is a key process both for the development of precipitation ( jiusto and weickmann 1973 ) , and in terms of the effect such clouds have on climate ( houghton 2001 ) . in this work ,
we use radar observations of deep cirrus to study the growth of ice particles as they sediment through the cloud .
vertically - pointing measurements of radar reflectivity @xmath1 and doppler velocity @xmath2 were made using the 35 ghz ( 8.6 mm ) ` copernicus ' radar at the chilbolton observatory in southern england . at this wavelength ,
the overwhelming majority of cirrus - sized particles are within the rayleigh regime where the backscattered intensity is proportional to the square of particle mass @xmath3 : @xmath4 where @xmath5 is the density of solid ice and @xmath6 is the number density of particles with mass between @xmath3 and @xmath7 .
the dielectric factor @xmath8 contains the information about the shape and dielectric strength of the particles : for spherical ice particles @xmath9 and the permittivity of ice @xmath10 at millimetre wavelengths is approximately 3.15 ( jiang and wu 2004 ) .
the rayleigh scattering approximation at 35 ghz is accurate to within 10% for particles with a maximum dimension of 1 mm or less ( westbrook _ et al _ 2006 ) .
the doppler velocity is @xmath11 , where @xmath12 is the @xmath13-weighted average terminal velocity of the particles and @xmath14 is the vertical air motion .
we use these measurements to estimate a characteristic particle fall time @xmath0 , which we define as the time for which the ` average ' particle ( with terminal velocity @xmath12 ) has been falling .
note that the doppler velocity is weighted by the reflectivity making it sensitive to the larger ice particles , and so our average fall time will also be weighted toward these large particles .
taking the cloud top height @xmath15 to be the altitude at which there is no longer a detectable radar return , we calculate the fall time associated with height @xmath16 as : @xmath17 given this new measure , we are in a position to investigate the evolution of the ice particles , by studying the variation of reflectivity @xmath1 with increasing fall time @xmath0 .
the advantage of this method , as opposed to simply studying @xmath1 as a function of height , is that @xmath0 represents the physical time for which the average ice particle has been falling to reach a given height @xmath16 , allowing us to relate our results to theoretical models of ice particle growth .
note that we have implicitly assumed that the cloud is in a steady state , such that the properties of the ice particles at height @xmath16 do not change significantly over the length of time it takes a particle to fall from cloud top to cloud base ( which is between 45 minutes and 2 hours for the cases shown here ) .
essentially this means that the cloud does not evolve significantly on this time scale and is advecting as a rigid body across the radar beam .
we therefore apply our technique only to non - precipitating , well developed ice clouds where there is there is low wind shear .
our case study is a cirrus cloud observed over chilbolton on the @xmath18 of may 2004 .
the temperature at cloud top ( as forecast by the met office mesoscale model , cullen 1993 ) was approximately @xmath19c , and the cloud base was close to @xmath20c ; the average wind shear over the depth of the cloud was approximately @xmath21 .
measurements of reflectivity and doppler velocity were made and the time series of these observations is shown in figure [ cloud ] .
the radar gate length is 30 m ( illingworth _ et al _ 2006 ) .
the values of @xmath1 and @xmath2 are averages over periods of 30 seconds : in figure [ cloud]c we also show the standard deviation @xmath22 of the 1-s average doppler velocity over each 30-s period , to indicate the small - scale variability in @xmath2 .
this measure allows the level of turbulence in the cloud to be assessed ( bouniol _ et al _ 2003 ) .
figure [ zt ] shows four representative vertical profiles sampled from different portions of the cloud , indicated by the dashed lines on figure [ cloud ] .
ten consecutive 30-s profiles were averaged over a period of @xmath237 minutes in order to smooth out the variability caused by fall streaks in the data .
the highest detectable cloud pixel ( corresponding to @xmath24 dbz ) from the profile is taken as a measure of cloud top .
the fall time at each height bin is calculated from the doppler velocity profile as per equation [ avtime ] , and we plot @xmath1 as a function of @xmath0 . from figure
[ zt ] we see that reflectivity increases rapidly with fall time ( note the logarithmic dbz units ) , which we interpret as rapid growth of the ice particles .
this could potentially be occurring through a number of possible mechanisms : deposition of water vapour ; aggregation via differential sedimentation of the ice particles ; or collisions with supercooled drops ( riming ) . in section 4
we show that it is likely that aggregation is the dominant growth mechanism .
the increase in @xmath1 appears to be exponential to a good approximation , and occurs for between 2500 and 5000 seconds in the profiles shown here .
the slopes on the log scale vary between approximately @xmath25 @xmath26 @xmath27 and @xmath28 @xmath26 @xmath27 , presumably depending on how much ice is being produced at cloud top . after this time there is a sharp turn over in the @xmath29 curves , and we attribute this to evaporation of the particles near cloud base .
such evaporation often results in increased air turbulence for which the particles themselves act as tracers , resulting in large variability in the doppler velocity . in the earlier profiles ( 07:09 and 07:38 utc ) this was not evident ; however , in the later profiles ( 08:06 and 08:30 utc ) the higher ice water content and time - integrated evaporative cooling triggered convective overturning and turbulence , and this is reflected in our observations ( figures [ cloud ] and [ zt ] ) , which show a sudden increase in @xmath22 at approximately the same time as the turn over in @xmath29 . exponential growth has also been observed in a number of other cloud data sets , and four more example profiles from well developed non - precipitating ice clouds during april and may 2004 are shown in figure [ otherclouds ] .
this is an interesting feature of the data , and a robust one in the face of errors in @xmath15 : if @xmath29 is exponential , then even if we have underestimated the cloud top somewhat ( on account of the limited sensitivity of the radar ) , this will merely correspond to an offset in the fall time , and the exponential shape of @xmath29 is still preserved .
it is interesting to note that the transition from growth to evaporation is not always sharp as it is for the @xmath30 may profiles : we speculate that this may be the result of aggregation continuing to some extent within the evaporation layer .
here we show how the reflectivity @xmath1 is related to the average particle size . scaling or `
normalised ' forms for the size distributions of both liquid and ice particles have been proposed in a number of recent articles ( rain : testud _ et al _ 2001 , illingworth and blackman 2002 , lee _ et al _ 2005 ; ice : field and heymsfield 2003 , westbrook _
et al _ 2004a , b , delano _ et al _ 2005 ) .
the essence of these rescaling schemes is that the underlying shape of the distribution @xmath31 is the same throughout the vertical profile , but is rescaled as a function of the ( increasing ) average particle mass @xmath32 as the particles grow : @xmath33 where we have normalised by the ice water content iwc .
the universal function @xmath34 is dimensionless .
equation [ dseqn ] indicates that a single average particle mass @xmath32 is sufficient to characterise the evolution of the particle size distribution ( relative to the iwc or some other moment of the distribution ) , and this is key to our analysis .
an example of such a distribution is that assumed in the uk met office s unified model ( wilson and ballard 1999 ) .
mass @xmath3 and diameter @xmath35 are assumed to be in a power law relationship @xmath36 , with an exponential distribution for particle diameter : @xmath37 where @xmath38 .
a single bulk prognostic variable is used for the ice particle mixing ratio and @xmath39 is parameterised to decrease with increasing temperature to mimic particle growth .
the parameter @xmath40 is calculated from the predicted iwc and @xmath39 , and is interpreted as a reciprocal average diameter @xmath41 ( eg .
et al _ 1995 ) . within the framework ( [ dseqn ] ) above
, this distribution corresponds to : @xmath42^{-1}x^{(1-b)/b}\exp\left({-x^{1/b}}\right),\ ] ] where @xmath43 , and @xmath44 .
irrespective of what form is assumed for @xmath45 , a scaling relationship between different moments of the distribution may be found .
the @xmath46 moment of the mass distribution is given by : @xmath47 note that @xmath48 is a dimensionless constant .
similarly , the radar reflectivity ( [ zeqn ] ) is given by : @xmath49 combining these two equations we may relate @xmath1 to an arbitrary moment @xmath50 of the distribution : @xmath51 at this point we make a crucial assumption : that there is some moment of the distribution @xmath52 which is approximately constant through the vertical profile .
in the case where aggregation is the dominant growth mechanism with a fixed production of ice mass at cloud top , one would expect the mass flux density of ice @xmath53 to be constant .
mitchell ( 1996 ) indicated that a power law for ice particle fall speeds @xmath54 is a good approximation : @xmath55 , so for pure aggregation @xmath56 . similarly , where diffusional growth or riming is dominant , the total number flux of particles would be roughly constant and @xmath57 would be the conserved moment .
if this assumption holds then the bracketted expression @xmath58 in equation [ scale ] is fixed through the vertical profile , and @xmath59 .
given our observations of exponential @xmath29 and the predicted power law between @xmath1 and @xmath32 above , we conclude that the average particle mass is growing exponentially with fall time .
we offer a possible explaination for the exponential growth of ice particles described above .
aircraft observations have indicated that aggregation is often the dominant growth mechanism for particles larger than a few hundred microns in cirrus clouds ( field and heymsfield 2003 ) , and it is these large particles which dominate the radar reflectivity .
recently , westbrook _
et al _ ( 2004a , b ) modelled ice particle aggregation by considering a rate of close approach between pairs of ice particles with masses @xmath3 and @xmath60 : @xmath61 where @xmath54 and @xmath62 are the associated fall speed and maximum dimension .
particles were picked according to the rate above , and traced along possible trajectories to accurately sample the collision geometries of the non - spherical ice particles .
the fall speeds were prescribed in the vein of mitchell ( 1996 ) : @xmath63 where the adjustable parameter @xmath64 determines the drag regime ( inertial flow @xmath65 ; viscous flow @xmath66 ) .
one of the key results from these simulations was that the aggregates produced by the model had a power law relationship between mass and maximum dimension @xmath67 , where the exponent is determined purely by the drag regime : @xmath68 for @xmath69 .
this relation is also backed up by a theoretical argument based on a feedback between the aggregate geometry and collision rate ( westbrook _ et al _ 2004b ) . for large snowflakes , @xmath70 and @xmath71 , in good agreement with aircraft observations ( eg .
@xmath72 , brown and francis 1995 ; @xmath73 , heymsfield _ et al _ 2002 ) .
in this study we are interested in the average ice particle growth rate , which is determined through the scaling of the collision kernel ( [ kernel ] ) . given the above relationship between @xmath74 and @xmath75 , and equations [ kernel ] and [ mitch ] , we see that if one doubles the masses of the aggregating particles @xmath76 : @xmath77 where @xmath78 .
this parameter controls the scaling of the particle growth rates and as such controls the growth of the average particle mass .
van dongen and ernst ( 1985 ) have shown that the coagulation equation ( pruppacher and klett 1997 ) has solutions with the same scaling form as ( [ dseqn ] ) , and predicts that the average particle mass grows according to the differential equation : @xmath79 where @xmath80 is a constant . given our prediction of @xmath81 from the aggregation model :
the prediction from aggregation theory is that average particle mass grows exponentially with fall time , in agreement with our observations .
we note that the van dongen and ernst analysis is for cases where total mass is conserved : however given the observed scaling behaviour ( [ dseqn ] ) and a power law relationship between mass and fall speed , the case where mass flux density is conserved should yield the same result .
the growth of particles by diffusion of water vapour may also be described by a similar equation to ( [ dmdt ] ) . however in that case @xmath83 and @xmath84 , where @xmath85 is the ` capacitance ' per unit diameter , @xmath86 is the supersaturation with respect to ice , and the terms @xmath87 and @xmath88 depend on temperature @xmath89 and pressure @xmath90 ( pruppacher and klett 1997 ) . for a given set of conditions @xmath91 , the growth by deposition
would be expected to increase slower with particle size than for aggregation , taking a power law form @xmath92 . in real clouds
these conditions do not stay constant , and there is a correlation between increasing particle size and increased temperature and supersaturation , which could lead to a faster growth rate .
however , it would take a considerable conspiracy between these variables to obtain a constant exponential growth throughout such an extensive region of the cloud as is observed in our radar data .
it also seems extremely unlikely that this correlation would be the same for all five cirrus cases shown in figures 2 and 3 .
we note that there is a region of sub - exponential growth close to cloud top ( small @xmath0 ) in some of the profiles in figure 3 : we suggest that it is in this region , where the particles are small and falling slowly , that diffusional growth dominates .
it seems very unlikely that riming dominated the ice particle growth : a large number of supercooled drops throughout the depth of the cloud would be required for this to be the case .
given the cold temperatures in the cloud ( between @xmath19c and @xmath20c as discussed in section 2 ) , it is very unlikely that supercooled drops would persist on long enough time scales and in large enough quantities to dominate the growth over the 2.5 km or so for which we have observed @xmath29 to increase exponentially .
we therefore discount deposition and riming , and assert that our observations are an indicator that aggregation is the dominant growth mechanism for the ice particles in these clouds .
doppler radar measurements of cirrus cloud were used to study the evolution of the ice particles sedimenting through it .
the results indicate that in the cases studied the average ice particle mass grows exponentially with fall time , in agreement with the theoretical expectation for aggregation , and we believe that this is evidence that aggregation of ice crystals is the dominant growth mechanism for large particles in deep , well developed ice clouds .
vertical profiles of reflectivity in ice have been much studied in order to estimate rainrates at the ground .
fabry and zawadzki ( 1995 ) observed an approximately constant d(dbz)/d@xmath16 gradient , and used this to rule out deposition as a growth mechanism .
this may be linked to our cirrus observations ; however their results were near the melting layer , and @xmath1 was @xmath2320db higher . we have compared profiles of dbz-@xmath16 and dbz-@xmath0 for our cirrus cases and find that while the dbz-@xmath0 profiles are straight lines with a constant gradient , the dbz-@xmath16 profiles have an appreciable curve to them .
the fact that our analysis ` straightens ' these curved profiles is good evidence that our approach of using the doppler velocity to estimate @xmath0 from @xmath16 is an appropriate one , and that aggregation is controlling the distribution of large ice particles .
the constant @xmath80 described in the aggregation theory above is directly related to the mass flux density , so measurements of the dbz-@xmath0 slope may allow the derivation of this quantity , and the data could be combined with doppler velocity measurements to estimate the ice water content . however , the sticking efficency of the ice particles ( which we assume to be constant with particle size ) is also a factor in @xmath80 , and this is a parameter for which there are few reliable experimental estimates . for warmer , ` stickier ' ice crystals at temperatures above @xmath93c this may be more feasible since the sticking efficiency should be close to unity .
we have assumed the ice particles fall vertically . in reality
there is likely to be some horizontal shear , and this , combined with variability in ice production of the cloud - top generating cells results in visible fall streaks ( see fig .
size - sorting along the streaks ( bader _ et al _ 1987 ) is a potential source of error in our analysis ; however , by averaging the reflectivity profiles over @xmath237 minutes of data we have been able to ameliorate it considerably .
directions for future work are to make dual - wavelength radar measurements of cirrus in order to obtain a more direct estimate of particle size ( westbrook _ et al _ 2006 ) .
this would help to pin down the dominant growth mechanism , allowing us to study moments other than @xmath1 , and analyse whether @xmath56 ( aggregation ) or @xmath57 ( deposition , riming ) is the moment conserved through the cloud .
aircraft observations ( field _ et al _ 2005 ) have indicated a broadly exponential trend between @xmath1 and temperature - it would be valuable to combine simultaneous radar and aircraft measurements to see if the exponential growth in @xmath1 with @xmath0 is accompanied by exponential growth in @xmath32 and increased concentrations of aggregates .
also , further studies of other cirrus cases , both at chilbolton and other radar sites , could be of interest to see how widespread the observed exponential trend is .
this work was funded by the natural environment research council ( grant number ner / z/2003/00643 ) .
we are grateful to the staff at the cclrc chilbolton observatory , ufam and the eu cloudnet project ( www.cloud-net.org ) , grant number evk2 - 2000 - 00065 .
we are grateful to our two reviewers for their valuable comments and suggestions .
bouniol , d. , a. j. illingworth and r. j. hogan ( 2003 ) , deriving turbulent kinetic energy dissipation rate within clouds using ground based 94 ghz radar , _ proc .
31@xmath94 ams conf . on radar
meteorology _ , seattle , 192196 .
brown , p. r. a. , a. j. illingworth , a. j. heymsfield , g. m. mcfarquhar , k. a. browning and m. gosset ( 1995 ) , the role of spaceborne millimetre - wave radar in the global monitoring of ice cloud , _
j. appl . met .
_ , _ 34 _ 23462366 . delano , j. , a. protat , j. testud , d. bouniol , a. j. heymsfield , a. bansemer , p. r. a. brown and r. m. forbes ( 2005 ) , statistical properties of the normalised ice particle distribution , _ j. geophys .
_ , _ 110 _ , d10201 . field , p. r. , r. j. hogan , p. r. a. brown , a. j. illingworth , t. w. choularton and r. j. cotton ( 2005 ) , parametrization of ice - particle size distributions for mid - latitude stratiform cloud , _ q. j. r. meteorol
_ , _ 131 _ , 19972017 .
heymsfield , a. j. , s. lewis , a. bansemer , j. iaquinta , l. m. miloshevich , m. kajikawa , c. twohy and m. r. poellot ( 2002 ) , a general approach for deriving the properties of cirrus and stratiform ice particles , _
_ , _ 60 _ , 17951808 . illingworth , a. j. and t. m. blackman ( 2002 ) , the need to represent raindrop size spectra as normalized gamma distributions for the interpretation of polarization radar observations , _ j. appl . met .
_ , _ 41 _ , 286297 .
testud , j. s. , s. oury , r. a. black , p. amayenc and x. k. dou ( 2001 ) , the concept of `` normalized '' distribution to describe raindrop spectra : a tool for cloud physics and remote sensing , _ j. appl . met .
_ , _ 40 _ , 11181140 . four ` snapshot ' vertical profiles from the cirrus case , taken at 07:09 , 07:38 , 08:06 , and 08:30 utc .
each profile shown is the average of ten consecutive 30-s profiles .
top row is reflectivity in dbz as a function of characteristic fall time ( points ) .
the solid line is intended to guide the eye , and indicates an exponential growth in @xmath1 with @xmath0 .
bottom row is @xmath22 as a function of @xmath0 , which we use as an indicator of particle evaporation near cloud base.,width=624 ] | vertically pointing doppler radar has been used to study the evolution of ice particles as they sediment through a cirrus cloud .
the measured doppler fall speeds , together with radar - derived estimates for the altitude of cloud top , are used to estimate a characteristic fall time @xmath0 for the ` average ' ice particle . the change in radar reflectivity @xmath1
is studied as a function of @xmath0 , and is found to increase exponentially with fall time .
we use the idea of dynamically scaling particle size distributions to show that this behaviour implies exponential growth of the average particle size , and argue that this exponential growth is a signature of ice crystal aggregation . |
knowledge of the low luminosity members of the galaxy the red dwarfs , subdwarfs , brown dwarfs , and white dwarfs ( wds ) comes largely from the nearest representatives of each class because they are the easiest to study .
magnitude - limited surveys of these objects allow us to develop crucial nearby samples that reveal ground truths about stellar populations .
the most straightforward technique to confirm proximity , to amass population statistics , and to better constrain physical parameters is precision astrometry , specifically trigonometric parallax determinations such as those presented in this paper .
the subjects of this paper , wds , are perhaps the most reliable chronometers of galactic history available to astronomers .
they also are valuable tracers of galactic populations of different ages , e.g. , the thin disk , thick disk , and halo . in addition , given that @xmath295% of all stars will end their lives as wds , the amount of mass contained within wds as a population may already be substantial and will continue to increase , possibly comprising a significant fraction of the current ( and future ) missing mass in the galaxy .
p5.0inccc + + & & & + + yale parallax catalog & 104 & 91 & 1 + _ hipparcos _ & 5 & 5 & 2,3,4,5 + torino observatory parallax program & 2 & 2 & 6 + ducourant and collaborators & 1 & 1 & 7 + ctiopi & 20 & 20 & 8 + total & 132 & 119 & + + + + [ wdsample ] + here we report results from our effort to enrich the sample of wds within 25 pc by discovering new wds and measuring accurate trigonometric parallaxes for those not yet measured . in this paper
, we also model nearby wds to determine accurate physical parameters , as well as evaluate population statistics for wds in the solar neighborhood .
spectral signatures and photometric spectral energy distributions ( seds ) are reproduced remarkably well by model atmospheres for most wds . however ,
a few exceptional wds , particularly the coolest members , remain problematic . here we present accurate parallaxes for two very cool and low luminosity wds in the solar neighborhood to provide empirical test cases for advancements in atmospheric modeling .
previous trigonometric parallax efforts have catalogued a total of 112 wd systems within 25 pc , using no uncertainty constraints .
a wd system is defined as any single or multiple stellar system containing at least one wd . the majority of these systems are included in the yale parallax catalog @xcite with recent additions from the _ hipparcos _ space astrometry mission @xcite as well as ground - based parallax programs and companion searches ( see table [ wdsample ] ) .
a comprehensive table containing the current 25 pc wd sample has been compiled as part of the dense ( discovery and evaluation of nearby stellar embers ) project and can be found at http://www.denseproject.com . surprisingly , only three wd parallaxes within 25 pc have been measured and published since the _ hipparcos _ catalog was first released more than a decade ago @xcite . to ensure a reliable nearby wd sample ,
we have adopted the quality limit for inclusion into the sample that the trigonometric parallax error can not be larger than 10% of the parallax . at 25 pc ,
this limit amounts to an error of 4.0 mas .
given the @xmath22 mas or better precision of ground - based parallaxes , this limit is entirely reasonable for the 25 pc sample . applying this constraint ,
13 systems are eliminated from the 25 pc wd sample , thus setting the total number of wd systems with reliable parallaxes within 25 pc to 99 . with the north and south galactic poles labeled as `` ngp '' and `` sgp '' , respectively .
the thick line is the galactic plane with the galactic center and the galactic anticenter labeled as `` gc '' and `` gac '' , respectively .
the encircled dot is the direction of the apex of solar motion and the encircled cross is the direction of the antapex of solar motion.,scaledwidth=45.0% ] in order to gauge the degree of incompleteness for the 25 pc wd sample , we make two basic assumptions , ( 1 ) that the 10 pc wd sample is complete , and ( 2 ) that the local density of wds out to 25 pc is constant . as is evident in figure [ incomplete ] , just over one - third of the wd systems expected within 25 pc have trigonometric parallaxes placing them within that volume .
a staggering @xmath263% are missing and this is merely a lower limit .
if additional wds are found within 10 pc ( such as the two presented here ) , the constant density curve shifts vertically upwards and increases the number of wds expected within 25 pc .
the small number of known wds within 10 pc ( 17 systems ) presents a fairly large uncertainty when extrapolating out to 25 pc solely because of small number statistics .
however , the fact remains that the sample is significantly incomplete . while there are a number of known wds likely within 25 pc
that do not yet have trigonometric parallaxes ( e.g. , * ? ? ?
* ) , there remains the need for a sizable sample of nearby , as yet undiscovered , wds to close the incompleteness gap .
the sky distribution of this sample of 99 wd systems is fairly homogeneous ( see figure [ skyplotnew ] ) .
however , if the sky is divided into four equal - area segments by declination , it is clear that the southern hemisphere is significantly undersampled in terms of nearby wd systems as shown in table [ skydist ] there are 63 systems in the north vs. 36 systems in the south prior to the inclusion of the results presented here .
in fact , the southernmost region is @xmath250% complete when compared to the northernmost region and , given that the 25 pc wd sample in the northernmost region is likely not complete , this is an upper limit .
thus , the southern hemisphere provides rich hunting grounds to identify new nearby wds for which our southern hemisphere parallax program , the cerro tololo inter - american observatory parallax investigation ( ctiopi ) , is ideally suited .
in general , there is an observational bias in that higher proper motion objects are scrutinized first because they are more likely to have larger parallaxes .
thus , there may exist a substantial number of wds with low proper motions that remains to be discovered , which may comprise a significant fraction of missing nearby wd systems . in order to evaluate empirically whether this proposition is plausible , we compared this sample to a subset of the _ hipparcos _ catalog .
our intention was to evaluate the proper motion distribution of a complete sample within 25 pc .
given that this catalog is complete to @xmath3 7.3 - 9.0 , we targeted all stars within 25 pc later than a0 ( assuming @xmath4 0.6 , * ? ? ?
* ) down to an apparent v magnitude of 7.5 ( @xmath4 5.5 , hence late g type dwarfs ) .
a constraint was used to remove evolved stars from the sample by defining a line @xmath21 mag above the locus of the main sequence in the h - r diagram and excluding those objects with brighter absolute magnitudes .
the resulting sift included a total of 325 stars , of which 51 ( @xmath216% ) had proper motions less than 0@xmath515 yr@xmath6 , which is the detection limit of @xcite and well below the detection limits of other proper motion surveys.1 yr@xmath6 ; however , they employed color constraints in search of late - type dwarfs that would eliminate wds from their sample . ]
thus , it seems the 25 pc wd sample is deficient in members with low proper motions because only 5 of the 99 systems known before this paper have @xmath7 0@xmath515 yr@xmath6 .
age effects , leading to larger velocity dispersions for older populations ( e.g. , wds ) , are not taken into account that could potentially alter these statistics , but likely not by the factor of roughly three necessary to resolve the discrepancy .
p1.8incc + + & & + & & + + @xmath890@xmath9 to @xmath830@xmath9 & 30 & 0 + @xmath830@xmath9 to @xmath800@xmath9 & 33 & 0 + total ( north ) & + @xmath1000@xmath9 to @xmath1030@xmath9 & 21 & 7 + @xmath1030@xmath9 to @xmath1090@xmath9 & 15 & 13 + total ( south ) & + + [ skydist ] in addition to measuring the first trigonometric parallaxes for several nearby wds , we are conducting an effort known as the astrometric search for planets encircling nearby stars ( aspens , * ? ? ?
the aspens effort monitors most red dwarfs within 10 pc and white dwarfs within 15 pc in the southern hemisphere to search for astrometric signals indicative of unseen companions .
the data are acquired in exactly the same manner as the parallax data except over longer time spans and perhaps with increased cadence .
in addition to probing for astrometric wobbles , these data allow us to redetermine trigonometric parallaxes for nearby wds that are often significantly more accurate than were previously available .
ironically , while the southern hemisphere is undersampled with respect to the 25 pc wd sample , the majority of 10 pc wd systems are in the south ( 11 of 17 ) .
thus , we present updated parallaxes for seven previously known wds within 10 pc . p0.75inlrrrcccrrrrrrrrl + + & & & & & & & & & & & & & & & & + & & & & & & & & & & & & & & & & + & & & & & & & & & & & & & & & & + + + + + 0038@xmath10226 & lhs 1126 & 14.50 & 14.08 & 13.66 & 3 & @xmath11 & 0.006 & 21 & 90 & 13.34 & 0.03 & 13.48 & 0.03 & 13.74 & 0.04 & + 0121@xmath10429 & lhs 1243 & 14.83 & 14.52 & 14.19 & 4 & @xmath11 & 0.005 & 11 & 60 & 13.86 & 0.02 & 13.63 & 0.04 & 13.53 & 0.04 & + 0141@xmath10675 & lhs 145 & 13.82 & 13.52 & 13.23 & 3 & @xmath12 & 0.007 & 29 & 166 & 12.87 & 0.02 & 12.66 & 0.03 & 12.58 & 0.03 & + 0419@xmath10487 & gj 2034 & 14.37 & 13.76 & 12.46 & 3 & @xmath11 & 0.010 & 12 & 64 & 10.72 & 0.02 & 10.15 & 0.02 & 9.85 & 0.03 & + 0628@xmath10020 & lp 600@xmath1042 & 15.32 & 15.06 & 14.75 & 3 & @xmath13 & [ 0.040 ] & 15 & 70 & & & & & & & + & lp 600@xmath1043 & 15.50 & 14.13 & 12.41 & 3 & @xmath13 & 0.007 & 15 & 74 & 10.73 & 0.03 & 10.14 & 0.03 & 9.86 & 0.02 & + 0751@xmath10252 & scr 0753@xmath102524 & 16.27 & 15.78 & 15.31 & 4 & @xmath11 & [ 0.029 ] & 13 & 45 & 14.75 & 0.03 & 14.47 & 0.03 & 14.30 & 0.09 & + 0806@xmath10661 & l97@xmath103 & 13.73 & 13.66 & 13.61 & 4 & @xmath11 & 0.007 & 13 & 65 & 13.70 & 0.02 & 13.74 & 0.03 & 13.78 & 0.04 & + 0821@xmath10669 & scr 0821@xmath106703 & 15.34 & 14.82 & 14.32 & 3 & @xmath11 & 0.007 & 17 & 86 & 13.79 & 0.03 & 13.57 & 0.03 & 13.34 & 0.04 & + 1009@xmath10184 & wt 1759 & 15.44 & 15.18 & 14.91 & 3 &
@xmath13 & 0.007 & 18 & 77 & 14.68 & 0.04 & 14.52 & 0.06 & 14.31 & 0.07 & + 1036@xmath10204 & lhs 2293 & 16.24 & 15.54 & 15.34 & 3 & @xmath11 & 0.006 & 14 & 52 & 14.63 & 0.03 & 14.35 & 0.04 & 14.04 & 0.07 & + 1202@xmath10232 & lp 852@xmath107 & 12.80 & 12.66 & 12.52 & 3 & @xmath11 & 0.007 & 16 & 75 & 12.40 & 0.02 & 12.30 & 0.03 & 12.34 & 0.03 & + 1223@xmath10659 & gj 2092 & 14.02 & 13.82 & 13.62 & 3 & @xmath12 & [ 0.028 ] & 14 & 61 & 13.33 & 0.04 & 13.26 & 0.06 & 13.30 & 0.06 & + 1315@xmath10781 & l40@xmath10116 & 16.15 & 15.74 & 15.36 & 3 & @xmath11 & 0.009 & 15 & 63 & 14.89 & 0.04 & 14.67 & 0.08 & 14.58 & 0.12 & + 1436@xmath10781 & ltt 5814 & 16.10 & 15.81 & 15.48 & 3 & @xmath11 & 0.006 & 18 & 63 & 15.04 & 0.04 & 14.88 & 0.08 & 14.76 & 0.14 & + 2008@xmath10600 & scr 2012@xmath105956 & 15.84 & 15.40 & 14.99 & 4 & @xmath12 & 0.008 & 21 & 84 & 14.93 & 0.05 & 15.23 & 0.11 & 15.41 & null & + 2008@xmath10799 & scr 2016@xmath107945 & 16.35 & 15.96 & 15.57 & 4 & @xmath11 & 0.008 & 12 & 55 & 15.11 & 0.04 & 15.03 & 0.08 & 14.64 & 0.09 & + 2040@xmath10392 & l495@xmath1082 & 13.75 & 13.76 & 13.69 & 3 & @xmath11 & 0.019 & 16 & 67 & 13.78 & 0.02 & 13.82 & 0.03 & 13.81 & 0.05 & + 2138@xmath10332 & l570@xmath1026 & 14.48 & 14.31 & 14.16 & 4 & @xmath12 & 0.007 & 16 & 67 & 14.17 & 0.03 & 14.08 & 0.04 & 13.95 & 0.06 & + 2336@xmath10079 & gd 1212 & 13.28 & 13.27 & 13.24 & 4 & @xmath11 & 0.009 & 19 & 74 & 13.34 & 0.03 & 13.34 & 0.02 & 13.35 & 0.03 & + 2351@xmath10335 & lhs 4040 & 14.52 & 14.38 & 14.19 & 3 & @xmath13 & [ 0.043 ] & 13 & 62 & 13.99 & 0.11 & 13.86 & 0.25 & 13.73 & 0.11 & + & lhs 4039 & 13.46 & 12.33 & 10.86 & 3 & @xmath13 & 0.008 & 13 & 62 & 9.48 & 0.02 & 8.91 & 0.02 & 8.61 & 0.02 & + + + + + 0928@xmath10713 & l64@xmath1040 & 15.11 & 14.97 & 14.83 & 3 & @xmath11 & 0.006 & 16 & 66 & 14.77 & 0.03 & 14.69 & 0.06 & 14.68 & 0.09 & + 1647@xmath10327 & lhs 3245 & 16.20 & 15.85 & 15.49 & 3 & @xmath11 & 0.009 & 12 & 46 & 15.15 & 0.05 & 14.82 & 0.08 & 14.76 & 0.11 & + 2007@xmath10219 & gj 781.3 & 14.40 & 14.33 & 14.25 & 3 & @xmath12 & 0.007 & 32 & 146 & 14.19 & 0.02 & 14.20 & 0.04 & 14.26 & 0.08 & + + + + + 0552@xmath10041 & lhs 32 & 14.47 & 13.99 & 13.51 & 3 & @xmath11 & 0.006 & 23 & 156 & 13.05 & 0.03 & 12.86 & 0.03 & 12.78 & 0.03 & + 0738@xmath10172 & lhs 235 & 13.06 & 12.89 & 12.72 & 4 & @xmath13 & 0.009 & 15 & 92 & 12.65 & 0.02 & 12.61 & 0.03 & 12.58 & 0.04 & + & lhs 234 & 16.69 & 14.69 & 12.41 & 4 & @xmath13 & 0.006 & 15 & 92 & 10.16 & 0.02 & 9.63 & 0.02 & 9.29 & 0.02 & + 0752@xmath10676 & lhs 34 & 13.96 & 13.58 & 13.20 & 3 & @xmath11 & 0.006 & 12 & 70 & 12.73 & 0.02 & 12.48 & 0.03 & 12.36 & 0.02 & + 0839@xmath10327 & lhs 253 & 11.86 & 11.77 & 11.65 & 3 & @xmath12 & 0.007 & 16 & 94 & 11.58 & 0.03 & 11.54 & 0.03 & 11.55 & 0.03 & + 1142@xmath10645 & lhs 43 & 11.50 & 11.34 & 11.20 & 3 & @xmath12 & 0.007 & 28 & 173 & 11.18 & 0.01 & 11.13 & 0.04 & 11.10 & 0.03 & + 2251@xmath10070 & lhs 69 & 15.70 & 15.11 & 14.56 & 3 & @xmath11 & 0.007 & 16 & 83 & 14.01 & 0.03 & 13.69 & 0.04 & 13.55 & 0.05 & + 2359@xmath10434 & lhs 1005 & 12.97 & 12.82 & 12.66 & 3 & @xmath11 & 0.007 & 12 & 87 & 12.60 & 0.03 & 12.43 & 0.02 & 12.45 & 0.02 & + + [ photometry ] +
observations have been collected during the ongoing ctiopi program that began in august 1999 .
ctiopi was conducted first as an noao surveys program through january 2003 using both the 0.9 m and the 1.5 m @xcite telescopes and has since operated as part of the smarts ( small and moderate aperture research telescope system ) consortium using the 0.9 m telescope @xcite .
the data , results , and procedures presented here correspond to the 0.9 m telescope program . on average ,
@xmath280 nights per year have been allocated to ctiopi observations . the standard ccd setup for ctiopi observations ( both photometric and astrometric ) utilizes only the central quarter of the 2048 @xmath14 2046 tektronix ccd camera with 0@xmath5401 pixel@xmath6 , yielding a 6@xmath158 square field of view .
the tek 2 @xmath16 , @xmath17 , and @xmath18 are 5475 , 6425 , and 8075 , respectively .
the tek 2 @xmath19 filter cracked in 2005 march and was replaced by the very similar tek 1 @xmath19 filter .
see @xmath20 [ sub : cracked ] for a discussion on the impact this switch has on the data . ]
( hereafter without the subscripts ) filter set was used to carry out the observations .
photometric observations have been collected since the inception of ctiopi during scheduled observing runs when sky conditions were photometric .
standard stars from @xcite and @xcite were taken nightly through a range of airmasses to calibrate fluxes to the johnson - kron - cousins system and to calculate extinction corrections .
bias subtraction and flat - fielding ( using calibration frames taken nightly ) were performed using standard iraf packages .
an aperture of 14@xmath21 diameter was used when possible ( consistent with * ? ? ?
* ) to determine stellar flux .
cosmic rays within this aperture were removed before flux extraction .
aperture corrections were applied when neighboring sources fell within the adopted aperture . in these cases ,
the largest aperture that did not include flux from the contaminating source was used and ranged from 4@xmath21 to 12@xmath21 in diameter .
total uncertainties ( including internal night - to - night variations as well as external fits to the standard stars ) in the optical photometry are @xmath22 0.03 mag in each filter @xcite . in the cases of wd 0628@xmath10020 and wd 2351@xmath10335 , the primaries ( red dwarfs ) were a significant contaminant ( especially in @xmath13 ) so that aperture photometry alone was not possible . instead , a psf fit using interactive data language ( idl ) package _
mpfit2dpeak _ was generated for the primaries and then subtracted from the images .
aperture photometry was then performed on the subtracted images .
the same procedure was performed for extracting photometry for the primaries ( i.e. , the wd secondaries were psf fit and removed ) .
our photometric results are given in table [ photometry ] , which is divided into three samples based on the trigonometric parallaxes presented here : ( 1 ) new 25 pc wd members , ( 2 ) wds beyond 25 pc , and ( 3 ) known 10 pc aspens targets .
companions to wds ( i.e. , lp600@xmath1043 , lhs 234 and lhs 4039 ) for which photometric analyses were performed are included in the table just below their wd companions .
multi - epoch optical @xmath23 photometry [ columns ( 3)-(5 ) ] as well as near - infrared @xmath24 photometry [ columns ( 11)-(16 ) ] extracted from the 2mass database are listed in table [ photometry ] .
we performed a photometric variability analysis of the parallax target ( hereafter referred to as the `` pi star '' ) relative to the reference stars using the parallax data taken in a single filter , as outlined in @xcite .
columns ( 6)-(10 ) list the number of different nights the object was observed for photometry , the parallax filter , the standard deviation of the pi star s magnitude in that filter from the variability analysis , the number of nights parallax data were taken , and the total number of parallax frames used in the variability analysis . in general , objects with standard deviation values larger than 0.02 mag are considered variable , those with values between 0.01 - 0.02 mag are likely variable at a few percent level , and those less than 0.01 mag are `` steady '' .
there are four cases where the standard deviations are larger than 0.02 mag ( wd 0628@xmath10020 , wd 0751@xmath10252 , wd 1223@xmath10659 , and wd 2351@xmath10335 ) and in all four cases , there are contaminating sources within a few arcseconds .
thus , the large standard deviations are due to varying degrees of contamination within the photometric aperture depending on seeing conditions rather than intrinsic variability .
two objects show standard deviations between 0.01 - 0.02 mag ( wd 0419@xmath10487 and wd 2040@xmath10392 ) .
wd 0419@xmath10487 has an unresolved red dwarf companion in a short period orbit ( see @xmath20 [ comments ] ) so that mild variability might be expected .
wd 2040@xmath10392 is a new pulsating zz ceti wd ( see @xmath20 [ comments ] ) .
a complete discussion of parallax data acquisition and reduction techniques can be found in @xcite .
briefly , once an object ( or system ) is selected to be observed for a trigonometric parallax determination , the object needs to be `` set up '' .
this process consists of selecting a telescope pointing as well as a filter bandpass ( @xmath23 ) with which all subsequent astrometry observations will be taken .
the telescope pointing is selected so that a fairly homogeneous distribution of reference stars encircle the pi star , do not reside near bad columns , and are as close as possible to the pi star rather than near the edges of the ccd .
also , every effort is made to place the pi star as close to the center of the ccd as possible .
of course , compromises are made in cases of sparse fields .
once sufficient data have been collected ( criteria that define a definitive parallax are outlined in @xmath20 [ sub : criteria ] ) , each frame is inspected and poor quality frames ( e.g. , bad seeing , telescope guiding problems ) are discarded .
centroids for the reference field and pi star are extracted using _ sextractor _ .
the centroids are corrected for differential color refraction ( dcr ) based on the color of each reference star and pi star .
typically , this correction shifts the stars positions by no more than a few mas .
automatic quality control constraints on ellipticity , elongation , and full width at have maximum ( fwhm ) eliminate frames and individual stars ( reference and pi ) of poor quality not noticed during manual inspection .
a fundamental trail plate is chosen as a reference plate and rotated based on a comparison with either the guide star catalog 2.2 or 2mass ( used by default except when there are too few stars in the pi star field that are catalogued in the 2mass database ) . all star centroids on other plates are recalibrated to account for different scaling in both the x and y directions , as well as the different amounts of translation and rotation .
a least - squares reduction via the gaussfit program @xcite is performed , assuming the reference star grid has @xmath25 0 and @xmath26 0 , where @xmath27 and @xmath28 are parallax and proper motion , respectively .
after verification of a good reference field ( e.g. , no reference stars within 100 pc , no problematic reference stars because of unresolved multiplicity ) , the final reduction produces a relative trigonometric parallax for the pi star .
even the reference stars trace out small parallax ellipses because these stars are not infinitely far away .
thus , a correction to absolute parallax must be performed .
this is accomplished in one of ( at least ) three different ways , ( 1 ) using a model of the galaxy for the disk and halo , ( 2 ) spectroscopic parallaxes for each of the reference stars , or ( 3 ) photometric parallaxes for each of the reference stars . because accurate photometry is needed for dcr correction and
is already available , we use photometric parallaxes for the reference stars to correct to absolute parallax using the ccd distance relations of @xcite .
these relations assume the reference stars are single main - sequence dwarfs and do not take into account contamination from evolved stars , unresolved double stars , or reddening .
we identify and remove outlying points estimated to be within @xmath2100 pc due either to evolved stars or unresolved double stars whose distances are underestimated .
given that all of the wds presented here are relatively nearby , pi star reddening is negligible .
however , the distant reference stars can be reddened and appear to be closer than they truly are .
this is the case for two wd fields ( wd 1223@xmath10659 and wd 1647@xmath10327 ) for which we have adopted an average correction to absolute of 1.2 mas ( based on all other wd fields presented here ) with a conservative error of 0.3 mas ( consistent with the largest absolute correction errors presented here ) . during the course of ctiopi , hundreds of parallaxes ( both preliminary and definitive ) have been reduced .
based on our experience with the `` evolution '' of the parallaxes as more data are added , we have set limits that define when a reduction is sufficient to be deemed definitive and hence publishable . in order to accurately decouple the parallactic and proper motions in the final astrometric solution , observations must span at least two years and adequately sample the parallax ellipse ( including high parallax factor observations ) .
also , there must be a balance between negative parallax factor observations and positive parallax factor observations and typically at least 20 good frames of each are required .
at least two and ideally three @xmath23 photometry observations per pi star are necessary to ensure accurate photometry needed to correct for dcr as well as to absolute parallax .
finally , the parallax error must be less than 3.0 mas .
this constraint was set early on during the program and with the use of the new centroiding algorithm ( see next section ) is typically an easy requirement to meet .
beginning with _ sextractor _ version 2.4 , the authors implemented windowed positional parameters .
these serve to alleviate a number of the inherent inaccuracies of the isophotal positional parameters used for centroiding in previous versions of _ sextractor _ ( see _ sextractor _
v2.0 user s guide for a complete discussion ) .
the authors analogize the positional accuracy of the gaussian - weighted two - dimensional windowed centroids to that offered by point - spread - function ( psf ) fitting .
after extensive testing on dozens of pi stars from the ctiopi program , we found that windowed centroids were clearly superior to the isophotal centroids used in our previous reductions , including those published in @xcite and @xcite from the 0.9 m . on average ,
parallax errors were reduced by @xmath250% with no appreciable change in the parallax ( when pi star contamination was not present ) using the same data sets in both reductions . in extreme cases , particularly in crowded fields , parallax errors were reduced by factors of two or more .
this was the case with the parallax reduction for wd 0751@xmath10252 , which had a contaminating source within 1@xmath50 . using the isophotal centroids , the preliminary parallax determination was 68.6 @xmath22 3.5 mas with clear trends in the residuals indicative of contamination .
using the windowed centroids and the same data set , the parallax determination was 55.8 @xmath22 1.0 mas with residuals that hovered near zero throughout the data 0.95 mas presented here , hence the slightly different value in table [ astrometry ] . ] . as an additional check
, this wd was discovered to be a common proper motion companion to ltt 2976 @xcite , which has a _ hipparcos _ parallax of 51.52 @xmath22 1.46 mas @xcite . while this value deviates by slightly more than 2@xmath29 of the value determined for wd 0751@xmath10252 , it is in far better agreement than the result using the isophotal centroids .
we obtain an average parallax error of 1.10 mas for the sample presented here using windowed centroids ( excluding wd 0628@xmath10020 and wd 2351@xmath10335 because the observations were optimized for their red dwarf companions and as such , the exposure times were shorter and the wd components were underexposed ) .
487 is discussed in @xmath20 [ comments].,scaledwidth=45.0% ] in early 2005 , the standard tek 2 @xmath12 filter that was included in the @xmath23 filter set used for parallax observations cracked .
it was replaced with another @xmath12 filter that had a very similar transmission profile . upon reduction of parallaxes including data taken in mid-2005 , the targets that were observed in the @xmath12 filter appeared to show subtle arches in the residuals once parallax and proper motion were fit and removed .
it became more evident as additional data were collected .
now that we have @xmath23 years of data after the filter switch , we are able to see its effect as a `` dip '' in the pi star residuals of a few mas for data taken in mid-2005 .
we have performed extensive tests and have concluded that inclusion of both @xmath12 filters data in the final parallax reduction is beneficial for constraining the parallax provided there are sufficient data that span at least @xmath21 - 2 years after the filter switch .
the major drawback is that any astrometric perturbations from unseen companions will likely be missed ( unless the photocentric shift is large ) because the residuals are contaminated . given that the majority of wd parallaxes presented here are taken in @xmath11 ( 19 vs. 7 at @xmath12 and 4 at @xmath13 ) the effect is minimal for this sample .
astrometric results for the wd systems ( including companions for which astrometric analyses were performed ) are listed in table [ astrometry ] . columns ( 4)-(9 ) list the filter used for parallax observations , the number of seasons the pi star was observed , total number of frames used in the parallax reduction , the time coverage and length of the parallax data , and the number of reference stars used .
the c in column ( 5 ) signifies that the observations were continuous throughout every season within the time coverage .
the s signifies that observations were scattered such that there is at least one season with only one night s data ( or no data for an entire season ) .
columns ( 10)-(12 ) list the relative parallax , correction to absolute , and the absolute parallax .
the proper motions and position angles quoted in columns ( 13 ) and ( 14 ) are those measured with respect to the reference field ( i.e. , relative , not corrected to a non - rotating frame of reference ) .
the tangential velocities quoted in column ( 15 ) are not corrected for solar motion . 645 , vs. the recently updated _ hipparcos _ parallax ( hip , * ? ? ?
* ) .,scaledwidth=45.0% ] as can be seen in the observational hertzsprung - russel ( h - r ) diagram in figure [ hrdiag ] , all but one of the observed wds lie within the realm of known wds .
the only exception is wd 0419@xmath10487 for which an unresolved red dwarf companion significantly contaminates the optical and near - ir photometry ( see @xmath20 [ comments ] ) . in figure
[ picomp ] , ctiopi parallaxes for wds previously known to be within 10 pc ( aspens targets ) are compared to their parallaxes from the yale parallax catalog ( ypc , * ? ? ?
* ) and , in one case ( wd 1142@xmath10645 ) , with the _ hipparcos _ parallax @xcite . agreements are reasonably good ( within 2@xmath29 ) , especially considering that all of the values from the ypc were obtained using photographic plates . of
the new 25 pc wd members from this effort , the majority tend to have proper motions on the lower end of the distribution ( see figure [ mupitan]_a _ ) .
based on the discussion in @xmath20 [ census ] , this trend is to be expected and likely represents the realm where the majority of nearby wds as yet undiscovered will be found .
for instance , @xmath290% of the 25 pc wd sample members have proper motions greater than wd 1202@xmath10232 with @xmath30 0@xmath5246 yr@xmath6 yet it is now the 25th nearest wd system ( 10.83 @xmath22 0.11 pc ) , and wd 2336@xmath10079 has the lowest proper motion of the sample with @xmath30 0@xmath5034 yr@xmath6 . while none of the new 25 pc members are within 8 pc , five were found between 8 and 15 pc ( see figure [ mupitan]_b _ ) .
this includes one object with a poorly constrained previous parallax ( wd 0038@xmath10226 , @xcite ) , two that have been suspected to be nearby for many years but whose trigonometric parallaxes were not determined until now ( wd 0141@xmath10675 and wd 1036@xmath10204 , @xcite ) , and two recent discoveries ( wd 0821@xmath10669 and wd 1202@xmath10232 , @xcite ) . of particular interest
is that the two recent discoveries are closer than the 13 pc limit set by @xcite from which they determine the local wd density .
while the most recent local wd density determination @xcite takes these two objects into account , they support the idea that more nearby wds may yet be found .
the distribution of tangential velocities for the new 25 pc members is unexceptional ( see figure [ mupitan]_c _ ) . only one object ( wd 2008@xmath10600 ) has a tangential velocity greater than 100 km sec@xmath6 ( see @xmath20 [ comments ] ) .
5 yr@xmath6 ( _ b _ ) distance binned by 2 pc and ( _ c _ ) tangential velocity binned by 20 km sec @xmath6 . in all three histograms
, the unshaded region represents the 99 previously known wds within 25 pc and the shaded region corresponds to the 20 new discoveries presented in this work . the dip in panel _
b _ at @xmath214 pc is an artifact of the binning.,scaledwidth=48.0% ]
atmospheric modeling procedures of the wds are identical to those presented in @xcite and references therein , with the exception that the trigonometric parallax constrains the surface gravity ( instead of assuming log @xmath31 8 as was done in that publication ) .
briefly , optical / near - ir magnitudes are converted into fluxes using the calibration of @xcite and compared to the seds predicted by the model atmosphere calculations .
the observed flux , @xmath32 , is related to the model flux by the equation @xmath33 where @xmath34 is the ratio of the radius of the star to its distance from earth , @xmath35 is the eddington flux ( dependent on @xmath0 , @xmath36 , and atmospheric composition ) properly averaged over the corresponding filter bandpass , and @xmath27 in this context is the mathematical constant ( elsewhere throughout this paper , @xmath27 refers to the trigonometric parallax angle ) .
our fitting technique relies on the nonlinear least - squares method of levenberg - marquardt @xcite , which is based on a steepest descent method .
the value of @xmath37 is taken as the sum over all bandpasses of the difference between both sides of equation ( 1 ) , weighted by the corresponding photometric uncertainties . only @xmath38 and [ @xmath39 are free parameters and the uncertainties of both parameters are obtained directly from the covariance matrix of the fit .
the main atmospheric constituent ( hydrogen or helium ) is determined by the presence of h@xmath40 from spectra published in the literature ( references listed in table [ parameters ] ) or by comparing fits obtained with both compositions .
we start with log @xmath41 and determine @xmath0 and [ @xmath42 , which combined with the distance @xmath43 obtained from the trigonometric parallax measurement yields directly the radius of the star @xmath11 .
the radius is then converted into mass using evolutionary models similar to those described in @xcite but with c / o cores , @xmath44 and @xmath45 ( representative of hydrogen - atmosphere wds ) , and @xmath46 and @xmath47 ( representative of helium - atmosphere wds).bergeron / coolingmodels/. ] in general , the log @xmath1 value obtained from the inferred mass and radius ( @xmath48 ) will be different from our initial guess of log @xmath41 , and the fitting procedure is thus repeated until an internal consistency in log @xmath1 is reached .
the parameter uncertainties are obtained by propagating the error of the trigonometric parallax measurements into the fitting procedure .
p0.75inr@@xmath22rccr@@xmath22rr@@xmath22rccr@@xmath22rr@@xmath22rr@@xmath22rr@@xmath22rc + + & & & & & & & & & & & & + & & & & & & & & & & & & + & & & & & & & & & & & & + + + + + 0038@xmath10226 & 110.42 & 1.17 & 2 & 1,2 & 5210 & 130 & 7.92 & 0.02 & he(+h ) & 6 & 0.52 & 0.01 & 14.72 & 0.04 & @xmath103.94 & 0.01 & 4.37 & 0.18 & + 0121@xmath10429 & 54.61 & 0.96 & 1 & 2 & & & & & & 13.52 & 0.05 & & & + 0141@xmath10675 & 102.80 & 0.85 & 1 & 2 & 6470 & 130 & 7.99 & 0.01 & h & 7 & 0.58 & 0.01 & 13.88 & 0.03 & @xmath103.58 & 0.01 & 1.81 & 0.04 & + 0419@xmath10487 & 49.68 & 1.34 & 1 & 2 & & & & & & 12.85 & 0.07 & & & + 0628@xmath10020 & 48.84 & 1.10 & 2 & 2,4 & 6846 & 383 & 8.06 & 0.04 & h & 8 & 0.63 & 0.02 & 13.76 & 0.06 & @xmath103.53 & 0.02 & 1.78 & 0.10 & + 0751@xmath10252 & 55.05 & 0.80 & 2 & 2,3 & 5160 & 100 & 7.97 & 0.02 & h & 9 & 0.56 & 0.01 & 14.97 & 0.04 & @xmath103.97 & 0.02 & 4.46 & 0.30 & + 0806@xmath10661 & 52.17 & 1.67 & 1 & 2 & 10250 & 70 & 8.06 & 0.05 & he(+c ) & 10 & 0.62 & 0.03 & 12.32 & 0.08 & @xmath102.83 & 0.03 & 0.67 & 0.04 & + 0821@xmath10669 & 93.89 & 1.04 & 1 & 2 & 5150 & 100 & 8.11 & 0.02 & h & 11 & 0.65 & 0.01 & 15.20 & 0.04 & @xmath104.06 & 0.01 & 6.22 & 0.16 & + 1009@xmath10184 & 55.55 & 0.85 & 2 & 2,3 & 5940 & 280 & 8.05 & 0.02 & he(+h , ca ) & 10 & 0.60 & 0.01 & 14.16 & 0.04 & @xmath103.78 & 0.02 & 2.97 & 0.18 & + 1036@xmath10204 & 70.00 & 0.66 & 1 & 2 & & & & & & 15.47 & 0.04 & & & + 1202@xmath10232 & 92.37 & 0.91 & 1 & 2 & 8590 & 170 & 7.90 & 0.02 & h & 12 & 0.54 & 0.01 & 12.63 & 0.04 & @xmath103.04 & 0.01 & 0.78 & 0.02 & + 1223@xmath10659 & 61.43 & 1.16 & 1 & 2 & 7660 & 220 & 7.81 & 0.03 & h & 13 & 0.49 & 0.02 & 12.96 & 0.05 & @xmath103.19 & 0.02 & 0.93 & 0.04 & + 1315@xmath10781 & 52.14 & 0.94 & 1 & 2 & 5730 & 160 & 8.18 & 0.03 & h & 11 & 0.70 & 0.02 & 14.74 & 0.05 & @xmath103.91 & 0.02 & 4.15 & 0.22 & + 1436@xmath10781 & 40.56 & 0.80 & 1 & 2 & 6270 & 200 & 8.08 & 0.03 & h & 11 & 0.64 & 0.02 & 14.14 & 0.05 & @xmath103.70 & 0.02 & 2.41 & 0.18 & + 2008@xmath10799 & 40.06 & 1.35 & 1 & 2 & 5800 & 160 & 7.97 & 0.06 & h & 11 & 0.57 & 0.03 & 14.36 & 0.08 & @xmath103.77 & 0.03 & 2.34 & 0.25 & + 2008@xmath10600 & 60.42 & 0.86 & 1 & 2 & 5080 & 220 & 7.89 & 0.03 & he(+h ) & 11 & 0.51 & 0.02 & 14.75 & 0.04 & @xmath103.97 & 0.02 & 4.55 & 0.27 & + 2040@xmath10392 & 44.18 & 0.97 & 1 & 2 & 10830 & 310 & 8.03 & 0.04 & h & 11 & 0.62 & 0.02 & 11.98 & 0.06 & @xmath102.71 & 0.02 & 0.51 & 0.02 & + 2138@xmath10332 & 64.00 & 1.41 & 1 & 2 & 7240 & 260 & 8.18 & 0.03 & he(+ca ) & 11 & 0.69 & 0.02 & 13.51 & 0.06 & @xmath103.51 & 0.02 & 1.96 & 0.12 & + 2336@xmath10079 & 62.72 & 1.70 & 1 & 2 & 11000 & 300 & 8.25 & 0.04 & h & 14 & 0.76 & 0.02 & 12.27 & 0.07 & @xmath102.81 & 0.03 & 0.69 & 0.04 & + 2351@xmath10335 & 43.74 & 1.43 & 2 & 2,4 & 8070 & 390 & 7.80 & 0.05 & h & 15 & 0.49 & 0.03 & 12.72 & 0.08 & @xmath103.10 & 0.03 & 0.81 & 0.05 & + + + + + 0928@xmath10713 & 38.44 & 0.64 & 1 & 2 & 8880 & 260 & 8.25 & 0.03 & h & 13 & 0.75 & 0.02 & 13.03 & 0.05 & @xmath103.19 & 0.02 & 1.18 & 0.04 & + 1647@xmath10327 & 37.03 & 1.18 & 1 & 2 & 6120 & 200 & 7.92 & 0.05 & h & 11 & 0.54 & 0.03 & 14.04 & 0.08 & @xmath103.64 & 0.03 & 1.87 & 0.14 & + 2007@xmath10219 & 38.22 & 0.94 & 1 & 2 & 9520 & 230 & 7.97 & 0.04 & h & 16 & 0.58 & 0.02 & 12.31 & 0.06 & @xmath102.90 & 0.03 & 0.66 & 0.03 & + + + + + 0552@xmath10041 & 155.97 & 0.78 & 2 & 1,2 & 5180 & 70 & 8.35 & 0.01 & he(+h , ca ) & 17 & 0.80 & 0.01 & 15.44 & 0.03 & @xmath104.20 & 0.01 & 6.82 & 0.02 & + 0738@xmath10172 & 109.43 & 0.56 & 3 & 1,2,4 & 7600 & 220 & 8.03 & 0.01 & he(+h , ca ) & 17 & 0.60 & 0.01 & 13.26 & 0.03 & @xmath103.34 & 0.01 & 1.41 & 0.02 & + 0752@xmath10676 & 126.62 & 1.32 & 2 & 1,2 & 5700 & 90 & 8.00 & 0.02 & h & 6 & 0.59 & 0.01 & 14.47 & 0.04 & @xmath103.81 & 0.01 & 2.65 & 0.10 & + 0839@xmath10327 & 113.59 & 1.93 & 2 & 1,2 & 9120 & 190 & 7.72 & 0.03 & h & 18 & 0.45 & 0.01 & 12.14 & 0.05 & @xmath102.84 & 0.02 & 0.55 & 0.02 & + 1142@xmath10645 & 216.12 & 1.09 & 3 & 1,2,5 & 7920 & 220 & 8.07 & 0.01 & he(+c ) & 19 & 0.62 & 0.01 & 13.17 & 0.03 & @xmath103.29 & 0.01 & 1.32 & 0.01 & + 2251@xmath10070 & 117.38 & 0.95 & 2 & 1,2 & 4000 & 200 & 7.92 & 0.02 & he(+h , ca ) & 17 & 0.52 & 0.01 & 16.05 & 0.03 & @xmath104.40 & 0.01 & 7.39 & 0.18 & + 2359@xmath10434 & 122.70 & 1.13 & 2 & 1,2 & 8530 & 160 & 8.39 & 0.01 & h & 16 & 0.85 & 0.01 & 13.40 & 0.04 & @xmath103.35 & 0.01 & 1.82 & 0.06 & + + + + + + + [ parameters ] + physical parameter determinations for the dq and dz wds are identical to the procedures outlined in @xcite .
briefly , the photometric sed provides a first estimate of the atmospheric parameters with an assumed value of metal abundances using solar abundance ratios .
the optical spectrum is fit to better constrain the metal abundances and to improve the atmospheric parameters from the photometric sed .
this procedure is iterated until a self - consistent photometric and spectroscopic solution is reached .
only two objects spectra are modeled here for the first time , wd 0806@xmath10661 and wd 1009@xmath10184 . for the remaining dq (
those with carbon ) and dz ( those with calcium ) stars , spectral modeling to obtain abundances was performed and presented in @xcite and @xcite .
the atmospheric abundances will not change with the inclusion of the parallaxes ; however , the surface gravities ( hence masses ) are sensitive to changes in distance and have been updated in table [ parameters ] . for wd 0806@xmath10661 , the optical spectrum shows no carbon absorption , yet it is classified as a dq based on ultraviolet ( uv ) spectra .
thus , the uv spectrum was used for fitting ( see @xmath20 [ comments ] ) . for dz stars ,
trace amounts of hydrogen not directly visible can be present in the atmosphere and affect the spectral profiles of the calcium absorption lines @xcite . in the case of wd 1009@xmath10184 ,
whose spectrum was obtained using the same telescope / instrument setup and reduction procedures as described in @xcite , the spectral fit including a log ( h / he ) @xmath493 better reproduced the calcium h & k lines than if no hydrogen were present ( see @xmath20 [ comments ] ) . in order to best
constrain the physical parameters for this sample , weighted mean parallaxes and errors are calculated for systems with previous parallax determinations as well as those that have common proper motion companions with previous / new parallax determinations .
the parallax values that are used to model the physical parameters are listed in table [ parameters ] [ column ( 2 ) ] as well as the number of individual parallaxes used in the weighted mean and corresponding references [ columns ( 3 ) and ( 4 ) ] . columns ( 5 ) and ( 6 ) list the effective temperatures and surface gravities as well as corresponding errors . columns ( 7 ) and ( 8) list the composition(s ) used in the atmospheric modeling ( with any secondary constituents listed in parentheses ) and the spectral type reference . columns ( 9)-(13 ) list the derived masses , absolute magnitudes , luminosities , wd ages ( not including main - sequence lifetimes ) , and any notes for the systems . while the errors listed for mass , luminosity , and age are formal errors , they are remarkably well constrained when accurate trigonometric parallaxes are available . * wd 0038@xmath10226 : * this wd has mild absorption bands similar to the carbon swan bands found in dq stars but the bands are shifted blueward .
initially , it was thought that these bands were actually the swan bands but were pressure - shifted because of increased pressures in cool he - rich wd atmospheres @xcite .
another explanation was that the features were caused by the hydrocarbon c@xmath50h @xcite .
the most recent possible explanations revisit the idea of pressure - shifted swan bands alone or in conjunction with swan bands produced by highly rotationally excited c@xmath50 @xcite .
additional theoretical investigations are necessary to better understand the properties of c@xmath50 in the high - pressure , high - temperature helium environment that a wd atmosphere would provide .
we present a significantly better parallax ( the previous parallax had an error @xmath51 10.3% ) , that confirms this object is within 10 pc and is the nearest known wd with the aforementioned spectral anomaly .
in addition , @xcite have shown that this object displays collision - induced absorption in the infrared and can be attributed to collisions of molecular hydrogen with helium . @xcite
have modeled the sed using a mixed h / he composition and arrived at a satisfactory fit .
the updated physical parameters found in table [ parameters ] were derived by an identical analysis ( including the use of their _ bvrijhk _ photometry ) except with the updated trigonometric parallax presented here . the helium abundance derived with the updated parallax
[ log ( he / h ) @xmath51 1.31 ] is less than that found by @xcite [ log ( he / h ) @xmath51 1.86 ] . * wd 0121@xmath10429 : * a recently discovered magnetic da that is thought to be an unresolved double degenerate with one component being a magnetic da and the other being a featureless dc @xcite .
the physical parameters are not listed in table [ parameters ] because any number of component masses and luminosities can reproduce the sed fit . * wd 0141@xmath10675 : * a da wd that was one of only two new 25 pc wd members to be within 10 pc . * wd 0419@xmath10487 : * a wd + red dwarf pre - cataclysmic eclipsing binary also known as with an orbital period of 7.3 hours for which @xcite derive masses and radii . contamination from the main - sequence component is significant both spectroscopically and photometrically so that no atmospheric modeling of the wd was possible ( it is the outlying point in the h - r diagram in figure [ hrdiag ] ) . * wd 0628@xmath10020 : * a known wd that has a red dwarf companion ( exact spectral type unknown ) with a separation of 4@xmath55 at 317.9@xmath9 .
parallax data were taken in @xmath13 to optimize observations for the red dwarf companion .
thus , the parallax for the wd is more uncertain because it is somewhat poorly exposed in our images .
the weighted mean parallax listed in table [ parameters ] , which represents our two measurements , should be taken as the distance to the system .
* wd 0738@xmath10172 : * a known nearby wd that has a m6.0v companion with a separation of 20@xmath56 at position angle 261.0@xmath9 .
because the companion is fainter in both @xmath12 and @xmath11 as well as redder , the parallax observations were taken in @xmath13 so that an independent parallax to the secondary red dwarf ( of comparable brightness to the primary at @xmath13 ) could also be obtained .
the parallax values for the pair are in excellent agreement ( see table [ astrometry ] ) but the proper motion values differ by several sigma of the formal errors . by comparing astrometric results of several distant wide binaries on ctiopi ( as yet unpublished ) , we find that the proper motion and position angle values agree to within 1 - 2@xmath29 .
thus , the formal errors are likely not significantly understated .
we make some basic assumptions about component masses and orientation ( i.e. , face - on orbit ) of the system to evaluate the plausibility of orbital motion to account for the discrepancy .
indeed , orbital motion during four years of observation of this nearby ( 9.14 @xmath22 0.05 pc ) system is likely the cause of the discrepancies in @xmath28 and position angle . * wd 0751@xmath10252 : * a recently discovered wd that is a common proper motion companion to ltt 2976 @xcite . the trigonometric parallax determined in this work for the wd ( 56.54 @xmath22 0.95 mas ) is in marginal agreement with the _ hipparcos _ parallax for the primary ( 51.52 @xmath22 1.46 mas , * ? ? ?
661 taken with _
( _ thin line _ ) as well as the model fit ( _ thick line _ ) and corresponding carbon abundance .
the poor fits to the two absorption lines at 1657 and 1930 are discussed in @xmath20 [ comments].,scaledwidth=48.0% ] * wd 0806@xmath10661 : * a dq wd that shows carbon features only in the uv ( in the optical , it appears as a featureless dc ) .
thus , a uv spectrum from the international ultraviolet explorer ( _ iue _ ) archive was included in the analysis .
first , a model that includes carbon with an abundance below the detectability limit in the optical [ e.g. , log ( c / he ) @xmath2 @xmath104 to @xmath106 ] was fit to the @xmath52 magnitudes .
however , the effective temperature and surface gravity derived from this model produced a poor fit to the uv spectrum .
then , the uv spectrum was fit independent of the photometry ( i.e. , @xmath0 and @xmath36 were allowed to vary ) .
this produced a much better fit with exception of two carbon lines at 1657 and 1930 ( see figure [ wd0806 ] ) .
the asymmetry in the observed spectrum is not reproduced in the model fit likely because of the failure of the impact approximation for the van der waals broadening used in the model . with the @xmath0 and @xmath53 fixed from the spectroscopic fit ,
the photometry was again fit to arrive at the final physical parameters found in table [ parameters ] , including an abundance of log ( c / he ) @xmath495.55 @xmath22 0.12 .
we note that the discrepancy in the temperatures derived from the uv spectrum and the optical / near - ir photometry is fairly small ( @xmath210,250 k vs. @xmath211,300 k ) , though significant .
it is possible that the models fail to address some component of the input physics , such as a missing opacity source that gives rise to this discrepancy between uv and optical effective temperature determinations . * wd
0821@xmath10669 : * a cool da wd that was uncovered during a trawl of the supercosmos sky survey ( sss ) database @xcite .
it is one of the oldest and nearest wds of the new 25 pc members 6.22 @xmath22 0.16 gyr at a distance of 10.65 @xmath22 0.12 pc .
thus , it is now the 23rd nearest wd system . * wd 1009@xmath10184 : * a dz wd that is difficult to model accurately because its effective temperature is near the point at which additional pressure effects become important and are not included in the dz models .
also , trace amounts of hydrogen may exist in the atmosphere but at levels too low to detect spectroscopically , yet affect the profile of the calcium absorption as discussed in @xcite .
our best fit using both photometry and spectroscopy produces a @xmath54 5940 @xmath22 280 k , including trace amounts of hydrogen [ log ( h / he ) @xmath493 ] and an abundance of log ( ca / he ) @xmath4910.37 @xmath22 0.20 ( see figure [ wd1009 ] ) .
however , it is likely that additional pressure effects have an impact , so the physical parameters should be considered preliminary estimates . 184 . the inset plot displays the spectrum ( _ thin line _ ) in the region to which the model ( _ thick line _ ) was fit ( assuming a slight hydrogen abundance of log ( h / he ) @xmath493 , see @xmath20 [ comments]).,scaledwidth=45.0% ] * wd 1036@xmath10204 : * a peculiar dq wd that has carbon swan absorption bands ; however , the absorption is significantly deeper in this object because of the presence of a large magnetic field @xcite .
again , no atmospheric modeling was possible but this object will serve as another important test case for model revisions in the future . *
wd 2008@xmath10600 : * a dc wd that was recently discovered to have a flux deficiency in the infrared because of collisions by molecular hydrogen with helium , similar to wd 0038@xmath10226 .
@xcite have modeled this object using a mixed h / he composition and a preliminary trigonometric parallax .
the updated physical parameters listed in table [ parameters ] replace the preliminary parallax with the more accurate one presented here , and we derive a helium abundance of log ( he / h ) @xmath51 2.60 . also , this object has the largest tangential velocity ( 112 km sec@xmath6 ) of the systems presented here . * wd 2040@xmath10392 : * a da wd that is listed in the mccook - sion white dwarf catalog @xcite but had no follow - up observations until @xcite obtained a photometric distance estimate of 23.1 @xmath22 4.0 pc ( in excellent agreement with the trigonometric parallax distance of 22.63 @xmath22 0.51 pc presented in this work ) .
in addition , as discussed in @xmath20 [ sec : phot ] , this object is variable at the @xmath22% level .
the physical parameters listed in table [ parameters ] ( i.e. , @xmath55 10,830 @xmath22 310 k , log @xmath31 8.03 @xmath22 0.04 , mass @xmath51 0.62 @xmath22 0.02 m@xmath56 , and absolute @xmath57 11.98 @xmath22 0.06 ) are entirely consistent with new pulsating zz ceti wds discovered by @xcite . for confirmation of zz ceti - type pulsations ,
data were acquired at the ctio 0.9 m using the same central quarter of the ccd as used for parallax data .
observations were taken in white light to maximize the signal for the target and eleven reference stars with a temporal resolution of @xmath21 minute .
relative aperture photometry was performed on the target and reference stars using an aperture diameter of 12@xmath21 .
the data span one hour , during which three cycles of a regular pulsation are clearly evident ( see figure [ zzceti ] ) , thus confirming the target is a zz ceti pulsator .
the fourier ( amplitude ) spectrum identifies the dominant period of @xmath2980 seconds with an amplitude of @xmath23% . 392 ( _ top panel _ ) normalized by its mean instrumental magnitude and the fourier ( amplitude ) spectrum ( _ bottom panel _
) identifying the dominant pulsational period.,scaledwidth=48.0% ] * wd 2251@xmath10070 : * a cool dz wd for which the atmospheric model appropriate for dzs fails to reproduce the observed spectrum ( see * ? ? ?
* ) likely because of additional pressure effects not accounted for in the model . also , this object s effective temperature is at the limit of the model grid .
physical parameters should be considered preliminary estimates . *
wd 2336@xmath10079 : * a da wd that @xcite recently discovered is a pulsating zz ceti wd . @xmath0 and @xmath36 listed in that publication are uncertain . with the high quality parallax presented here ,
the parameters are now better constrained and in reasonable agreement with @xcite . also , with a minuscule proper motion of 33.6 @xmath22 1.0 mas , this object is the slowest moving wd known in the 25 pc wd sample .
in fact , it was first cataloged by @xcite and labeled as a suspected wd simply based on its blue color .
indeed , this object has the largest effective temperature of all of the systems presented here .
thus , this object provides additional support for the possibility that a sample of nearby wds with little to no proper motions may exist , especially given that the majority of wds in a volume - limited sample are cooler such that their colors alone are unexceptional . *
wd 2351@xmath10335 : * a da wd that has both a m4v primary ( lhs 4039 ) and a recently discovered active m8.5v companion ( , @xcite ) .
the wd has a separation of 6@xmath55 at position angle 182.2@xmath9 from the primary and the m8.5v has a separation of 102@xmath58 at position angle 91.3@xmath9 from the primary .
parallax data were taken in @xmath13 to optimize observations for the m4v primary .
thus , the secondary wd was poorly exposed in the images giving rise to the largest parallax error of those presented here .
the weighted mean parallax listed in table [ parameters ] , which represents our two measurements , should be taken as the parallax of the system .
the tertiary m8.5v was unknown at the time parallax observations began ( mid-2003 ) so , while the object is visible in our frames , exposure times were not sufficient to obtain usable astrometry .
we have increased the number of wd systems with reliable trigonometric parallaxes within 25 pc from 99 to 119 ( 20% ) .
all are in the southern hemisphere ( see figure [ skyplotnew ] ) , thus , we are shrinking the disparity between the number of nearby wds in the two hemispheres illustrated in table [ skydist ] .
in fact , this effort has nearly doubled the number of 25 pc wds in the @xmath1030@xmath9 to @xmath1090@xmath9 quadrant of the sky . while 13 of the new 25 pc members have been known wds for many years , seven are recent wd discoveries @xcite .
trigonometric parallax determinations are underway for an additional @xmath230 wd systems estimated to be within 25 pc from those publications as well as others ( e.g. , * ? ? ?
* ) , and it is likely that we will continue to populate the 25 pc wd sample with reliable members . a number of interesting objects have been uncovered through the collection of parallax data presented here coupled with previous efforts data .
for instance , the parallax of wd 0121@xmath10429 places a considerable constraint on its mass of 0.41 m@xmath56 assuming a single wd . because this value is likely too small to have formed via single star evolution , the parallax gives additional weight to the hypothesis adopted from spectral analyses that this object is an unresolved double degenerate . also , a new
zz ceti pulsating wd was identified ( wd 2040@xmath10392 ) .
finally , the high quality parallaxes for two cool wds that display continuum absorption in the near - ir ( wd 0038@xmath10226 and wd 2008@xmath10600 ) show that trace amounts of hydrogen in a helium - dominated atmosphere is the most likely scenario to explain this poorly understood phenomenon . these objects and others like
it will serve as empirical checks and permit revisions to atmospheric models for the coolest wds .
in the course of collecting long time - series astrometric data for nearby wds , we hope to identify new systems that have astrometric perturbations from unseen companions .
in particular , identification of new double degenerate systems that are resolvable with high resolution astrometric instruments ( i.e. , _ hubble space telescope _ s fine guidance sensors ) will permit accurate dynamical mass determinations for two wds at once . to date , only three wd systems ( sirius b , , and , @xcite ) have dynamical masses known to better than 5% . in two of these cases ,
the largest uncertainties in the mass error budgets are the distance determinations even though all three systems are among the 50 nearest systems to the sun . also , these three systems serve as the only reliable empirical verifications of the theoretical wd mass - radius relation commonly used in wd modeling ( e.g. , wd ages via cooling models ) . thus , accurate trigonometric parallaxes of nearby wds are essential to finding additional systems whose masses can be well constrained . _
note added in manuscript_.two recent trigonometric parallax works were published during the review process of this manuscript that are relevant to the 25 pc wd sample . @xcite
present trigonometric parallaxes for 21 systems including four new wd systems within 25 pc ( wd 0423@xmath8044 , wd 0511@xmath8079 , wd 1309@xmath8853 the previous parallax had a parallax error larger than 10% , and wd 2047@xmath8372 ) .
@xcite present trigonometric parallaxes for 18 systems including two new wd systems within 25 pc ( wd 1814@xmath8134 and wd 2322@xmath8137 ) .
thus , the 25 pc wd sample contains 105 systems prior to this work and 125 systems including this work .
we would like to thank referee , john thorstensen , for his helpful comments and suggestions that made this publication more clear and concise .
the recons team at georgia state university wishes to thank the nsf ( grant ast 05 - 07711 ) , nasa s space interferometry mission , the space telescope science institute ( grant hst - go-10928.01-a ) , and gsu for their continued support in our study of nearby stars .
p. b. wishes to acknowledge the support of nserc canada as well as the the fund fqrnt ( qubec ) .
b. is a cottrell scholar of research corporation for science advancement .
e. c. and r. a. m. acknowledge support by the fondo nacional de investigacin cientfica y tecnolgica ( proyecto no . 1010137 , fondecyt ) and by the chilean centro de astrofsica fondap ( no .
15010003 ) .
we also thank the members of the smarts consortium , who enable the operations of the small telescopes at ctio , as well as the observer support at ctio , specifically edgardo cosgrove , arturo gomez , alberto miranda , and joselino vasquez .
this publication makes use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation .
gliese , w. , & jahrei , h. 1991 , on : the astronomical data center cd - rom : selected astronomical catalogs , vol .
i ; l.e .
brotzmann , s.e .
gesser ( eds . ) , nasa / astronomical data center , goddard space flight center , greenbelt , md , van leeuwen , f.2007 , hipparcos , the new reduction of the raw data . by floor van leeuwen , institute of astronomy , cambridge university , cambridge , uk
series : astrophysics and space science library , vol .
350 20 springer dordrecht + + & & & & & & & & & & & & & & & + & & & & & & & & & & & & & & & + & & & & & & & & & & & & & & & + + + + + & 00 41 26.03 & @xmath1022 21 02.3 & @xmath11 & 9s & 90 & [email protected] & 8.25 & 8 & 109.30 & 1.18 & 1.24 & 0.07 & 110.54 & 1.18 & 604.7 & 0.4 & 232.6 & 0.07 & 25.9 & + & 01 24 03.98 & @xmath1042 40 38.5 & @xmath11 & 5s & 60 & [email protected] & 3.90 & 6 & 54.07 & 0.96 & 0.54 & 0.03 & 54.61 & 0.96 & 594.1 & 0.7 & 151.0 & 0.12 & 51.6 & + & 01 43 00.98 & @xmath1067 18 30.3 & @xmath12 & 8c & 166 & [email protected] & 7.42 & 6 & 101.92 & 0.85 & 0.88 & 0.07 & 102.80 & 0.85 & 1079.7 & 0.4 & 199.0 & 0.04 & 47.8 & + & 04 21 05.56 & @xmath1048 39 07.1 & @xmath11 & 5s & 64 & [email protected] & 3.80 & 7 & 48.73 & 1.34 & 0.95 & 0.09 & 49.68 & 1.34 & 538.6 & 1.3 & 178.1 & 0.21 & 51.4 & + & 06 30 39.01 & @xmath1002 05 50.6 & @xmath13 & 5s & 70 & [email protected] & 3.88 & 6 & 43.64 & 1.73 & 2.87 & 0.31 & 46.51 & 1.76 & 205.6 & 1.8 & 214.5 & 0.97 & 21.0 & + & 06 30 38.80 & @xmath1002 05 54.0 & @xmath13 & 5s & 74 & [email protected] & 3.88 & 6 & 47.45 & 1.37 & 2.87 & 0.31 & 50.32 & 1.40 & 196.6 & 1.4 & 215.9 & 0.80 & 18.5 & + & 07 53 56.61 & @xmath1025 24 01.5 & @xmath11 & 4s & 45 & [email protected] & 2.67 & 11 & 56.01 & 0.95 & 0.53 & 0.03 & 56.54 & 0.95 & 362.2 & 1.3 & 304.7 & 0.40 & 30.4 & + & 08 06 53.76 & @xmath1066 18 16.6 & @xmath11 & 5s & 65 & [email protected] & 3.74 & 9 & 50.54 & 1.66 & 1.63 & 0.14 & 52.17 & 1.67 & 446.8 & 1.8 & 130.4 & 0.46 & 40.6 & + & 08 21 26.70 & @xmath1067 03 20.1 & @xmath11 & 6s & 86 & [email protected] & 4.95 & 11 & 92.99 & 1.04 & 0.90 & 0.07 & 93.89 & 1.04 & 762.3 & 0.7 & 329.5 & 0.10 & 38.5 & + & 10 12 01.88 & @xmath1018 43 33.2 & @xmath13 & 6s & 77 & [email protected] & 5.73 & 10 & 53.99 & 0.98 & 0.64 & 0.05 & 54.63 & 0.98 & 514.4 & 0.5 & 269.0 & 0.09 & 44.6 & + & 10 38 55.57 & @xmath1020 40 56.7 & @xmath11 & 4c & 52 & [email protected] & 3.13 & 8 & 69.32 & 0.66 & 0.68 & 0.04 & 70.00 & 0.66 & 610.0 & 0.6 & 334.0 & 0.11 & 41.3 & + & 12 05 26.66 & @xmath1023 33 12.1 & @xmath11 & 5s & 75 & [email protected] & 4.28 & 8 & 90.75 & 0.90 & 1.62 & 0.12 & 92.37 & 0.91 & 245.8 & 0.8 & 16.6 & 0.32 & 12.6 & + & 12 26 42.02 & @xmath1066 12 18.6 & @xmath12 & 5s & 61 & [email protected] & 3.97 & 10 & 60.33 & 1.12 & 1.20 & 0.30 & 61.53 & 1.16 & 185.8 & 0.9 & 186.9 & 0.42 & 14.3 & + & 13 19 25.63 & @xmath1078 23 28.3 & @xmath11 & 4c & 63 & [email protected] & 2.81 & 10 & 50.78 & 0.93 & 1.36 & 0.11 & 52.14 & 0.94 & 470.0 & 1.1 & 139.5 & 0.27 & 42.7 & + & 14 42 51.51 & @xmath1078 23 53.6 & @xmath11 & 5s & 63 & [email protected] & 4.54 & 12 & 39.80 & 0.80 & 0.76 & 0.05 & 40.56 & 0.80 & 409.6 & 0.7 & 275.1 & 0.16 & 47.9 & + & 20 12 31.75 & @xmath1059 56 51.5 & @xmath12 & 4s & 84 & [email protected] & 3.06 & 13 & 59.42 & 0.86 & 1.00 & 0.05 & 60.42 & 0.86 & 1427.6 & 1.0 & 166.1 & 0.07 & 112.0 & + & 20 16 49.74 & @xmath1079 45 53.0 & @xmath11 & 4s & 55 & [email protected] & 2.89 & 10 & 39.32 & 1.34 & 0.74 & 0.12 & 40.06 & 1.35 & 427.6 & 1.7 & 129.2 & 0.44 & 50.6 & + & 20 43 49.21 & @xmath1039 03 18.0 & @xmath11 & 5c & 67 & [email protected] & 4.21 & 11 & 43.13 & 0.96 & 1.05 & 0.15 & 44.18 & 0.97 & 339.3 & 0.7 & 182.2 & 0.19 & 36.4 & + & 21 41 57.56 & @xmath1033 00 29.8 & @xmath12 & 3c & 67 & [email protected] & 2.43 & 11 & 63.21 & 1.41 & 0.79 & 0.07 & 64.00 & 1.41 & 204.2 & 1.5 & 238.3 & 0.81 & 15.1 & + & 23 38 50.74 & @xmath1007 41 19.9 & @xmath11 & 5c & 74 & [email protected] & 4.31 & 8 & 61.67 & 1.70 & 1.05 & 0.10 & 62.72 & 1.70 & 33.6 & 1.0 & 126.6 & 3.22 & 2.5 & + & 23 54 01.14 & @xmath1033 16 30.3 & @xmath13 & 4c & 62 & [email protected] & 4.23 & 5 & 40.33 & 2.40 & 2.49 & 0.11 & 42.82 & 2.40 & 508.1 & 2.1 & 219.4 & 0.46 & 56.2 & + & 23 54 01.11 & @xmath1033 16 22.7 & @xmath13 & 4c & 62 & [email protected] & 4.23 & 5 & 41.75 & 1.78 & 2.49 & 0.11 & 44.24 & 1.78 & 515.4 & 1.5 & 218.0 & 0.33 & 55.2 & + + + + + & 09 29 07.97 & @xmath1071 33 58.8 & @xmath11 & 5c & 66 & [email protected] & 3.83 & 11 & 37.52 & 0.64 & 0.92 & 0.06 & 38.44 & 0.64 & 442.0 & 0.6 & 320.2 & 0.15 & 54.5 & + & 16 50 44.32 & @xmath1032 49 23.2 & @xmath11 & 4s & 46 & [email protected] & 2.88 & 11 & 35.93 & 1.14 & 1.20 & 0.30 & 37.13 & 1.18 & 501.3 & 1.2 & 193.2 & 0.24 & 64.2 & + & 20 10 17.51 & @xmath1021 46 45.6 & @xmath12 & 8s & 123 & [email protected] & 8.06 & 10 & 37.45 & 0.94 & 0.77 & 0.08 & 38.22 & 0.94 & 331.0 & 0.4 & 162.2 & 0.13 & 41.0 & + + + + + & 05 55 09.53 & @xmath1004 10 07.1 & @xmath11 & 5c & 156 & [email protected] & 4.19 & 9 & 153.78 & 0.75 & 2.35 & 0.38 & 156.13 & 0.84 & 2376.0 & 0.6 & 166.6 & 0.02 & 72.1 & + & 07 40 20.78 & @xmath1017 24 49.2 & @xmath13 & 5c & 92 & [email protected] & 4.18 & 11 & 108.73 & 0.80 & 1.06 & 0.09 & 109.79 & 0.81 & 1263.4 & 0.5 & 116.6 & 0.04 & 54.5 & + & 07 40 19.36 & @xmath1017 24 46.0 & @xmath13 & 5c & 92 & [email protected] & 4.18 & 11 & 108.81 & 0.81 & 1.06 & 0.09 & 109.87 & 0.82 & 1272.7 & 0.5 & 116.0 & 0.04 & 54.9 & + & 07 53 08.16 & @xmath1067 47 31.5 & @xmath11 & 5c & 70 & [email protected] & 4.19 & 13 & 125.14 & 1.33 & 1.11 & 0.15 & 126.25 & 1.34 & 2097.5 & 0.7 & 135.8 & 0.04 & 78.7 & + & 08 41 32.42 & @xmath1032 56 32.8 & @xmath12 & 5c & 94 & [email protected] & 4.34 & 8 & 112.24 & 1.97 & 1.39 & 0.10 & 113.63 & 1.97 & 1702.5 & 1.0 & 322.7 & 0.06 & 71.0 & + & 11 45 42.93 & @xmath1064 50 29.7 & @xmath12 & 9s & 173 & [email protected] & 8.23 & 10 & 214.16 & 1.24 & 1.64 & 0.19 & 215.80 & 1.25 & 2692.7 & 0.5 & 97.5 & 0.02 & 59.1 & + & 22 53 53.35 & @xmath1006 46 54.5 & @xmath11 & 5s & 83 & [email protected] & 4.37 & 8 & 115.57 & 0.96 & 1.49 & 0.14 & 117.06 & 0.97 & 2571.8 & 0.5 & 105.6 & 0.02 & 104.1 & + & 00 02 10.72 & @xmath1043 09 55.5 & @xmath11 & 6s & 92 & [email protected] & 4.87 & 7 & 121.02 & 1.11 & 1.25 & 0.19 & 122.27 & 1.13 & 887.8 & 0.8 & 138.4 & 0.10 & 34.4 & + + | we present accurate trigonometric parallaxes for 20 new members of the 25 pc white dwarf sample as part of the dense project ( discovery and evalution of nearby stellar embers , http://www.denseproject.com ) .
previously , there were a total of 112 white dwarf systems with trigonometric parallaxes placing them within 25 pc and of these , 99 have trigonometric parallaxes known to better than 10% .
thus , the 20 new members presented in this work represent a 20% increase in the number of white dwarfs accurately known to be within 25 pc .
in addition , we present updated parallaxes for seven known white dwarfs within 10 pc that have been observed as part of the aspens initiative ( astrometric search for planets encircling nearby stars ) to monitor nearby southern red and white dwarfs for astrometric perturbations from unseen companions . including a few white dwarf companions and white dwarfs beyond 25 pc , we present a total of 33 trigonometric parallaxes .
we perform atmospheric modeling for white dwarfs to determine physical parameters ( i.e. , @xmath0 , log @xmath1 , mass , and white dwarf age ) . finally , a new zz ceti pulsating white dwarf was identified and revised constraints are placed on two mixed h / he atmosphere cool white dwarfs that display continuum absorption in the near - infrared . |
a well - known explanation of the solar @xmath5 flux deficit @xcite is provided by flavor oscillations @xcite of neutrinos along their way from the sun @xmath6 to the earth @xmath7 . for two active neutrino states [ say ,
@xmath8 in the flavor basis and @xmath9 in the mass basis ] , the physics of solar @xmath4 oscillations is governed , at any given energy @xmath10 , by the mass - mixing parameters @xmath11 and @xmath12 in vacuum , , as far as @xmath12 is taken in the first quadrant @xmath13 $ ] . ] as well as by the electron density profile @xmath14 in matter @xcite .
different oscillation regimes can be identified in terms of three characteristics lengths , namely , the astronomical unit @xmath15 the oscillation length in vacuum @xmath16 and the refraction length in matter @xmath17 which is associated to the effective mixing angle @xmath18 @xcite , @xmath19 typical solutions to the solar neutrino problem ( see , e.g. , @xcite ) involve values of @xmath20 either in the so - called `` just - so '' ( js ) oscillation regime @xcite , characterized by @xmath21 or in the `` mykheyev - smirnov - wolfenstein '' ( msw ) oscillation regime @xcite , characterized by @xmath22 the two regimes correspond roughly to @xmath23 ev@xmath2/mev and to @xmath24 ev@xmath2/mev , respectively . for just - so oscillations , since @xmath25 ,
the effect of matter is basically to suppress the oscillation amplitude both in the sun and in the earth ( @xmath26 ) , so that ( coherent ) flavor oscillations take place only in vacuum , starting from the sun surface @xcite .
conversely , for msw oscillations , @xmath27 and flavor transitions are dominated by the detailed matter density profile , while the many oscillation cycles in vacuum @xmath28 are responsible for complete @xmath4 decoherence at the earth , once smearing effects are taken into account @xcite .
therefore , it is intuitively clear that in the intermediate range between ( [ js ] ) and ( [ msw ] ) , corresponding approximately to @xmath29 ev@xmath2/mev , the simple vacuum oscillation picture of the js regime becomes increasingly decoherent and affected by matter effects for increasing values of @xmath0 , leading to a hybrid regime that might be called of `` quasi - vacuum '' ( qv ) oscillations , characterized by @xmath30 in the past , quasi - vacuum effects on the oscillation amplitude and phase have been explicitly considered only in relatively few papers ( see , e.g. , @xcite ) as compared with the vast literature on solar neutrino oscillations , essentially because typical fits to solar @xmath4 rates allowed only marginal solutions in the range where qv effects are relevant .
however , more recent analyses appear to extend the former ranges of the js solutions _ upwards _
@xcite and of the msw solutions _ downwards _
@xcite in @xmath0 , making them eventually merge in the qv range @xcite , especially under generous assumptions about the experimental or theoretical @xmath4 flux uncertainties .
therefore , a fresh look at qv corrections seems warranted .
recently , a semianalytical approximation improving the familiar just - so formula in the qv regime was discussed in @xcite and , in more detail , in @xcite , where additional numerical checks were performed . in this work
we revisit the whole topic , by performing accurate numerical calculations which include the exact density profile in the sun and in the earth , within the reference mass - mixing ranges @xmath31 $ ] ev@xmath2/mev and @xmath32 $ ] .
is useful to chart the first two octants of the mixing angle range @xcite .
] we also discuss some approximations that can simplify the computing task in present applications .
we then apply such calculations to a global analysis of solar neutrino data in the range @xmath33 ev@xmath2 .
our paper is structured as follows .
the basic notation and the numerical techniques used in the calculations are introduced in sec .
[ s2 ] and [ s3 ] , respectively .
the effects of solar matter in the quasi - vacuum oscillation regime are discussed in sec .
[ s4 ] , where the results for true and exponential density profiles are compared .
earth matter effects are described in sec . [ s5 ] .
the decoherence of oscillations induced by energy ( and time ) integration is discussed in sec .
the basic results are summarized and organized in sec .
[ s7 ] , and then applied to a three - flavor oscillation analysis in sec .
section [ s9 ] concludes our work .
the @xmath4 propagation from the sun core to the detector at the earth can be interpreted as a `` double slit experiment , '' where the original @xmath5 can take two paths , corresponding to the intermediate transitions @xmath34 and to @xmath35 .
the global @xmath5 survival amplitude is then the sum of the amplitudes along the two paths , @xmath36 where the transition amplitudes from the sun production point to its surface @xmath37 , from the sun surface to the earth surface @xmath38 and from the earth surface to the detector @xmath39 have been explicitly factorized . the @xmath5 survival probability @xmath40 is then given by @xmath41 in general , the above amplitudes can be written as @xmath42 for the first path and as @xmath43 for the second path , where @xmath44 is the sun radius .
is relatively small ( @xmath45 ) , it is explicitly kept for later purposes .
the earth radius @xmath46 can instead be safely neglected ( @xmath47 ) . ] in the above equations , @xmath48 and @xmath49 are real numbers ( @xmath50 $ ] ) representing the transition probability @xmath51 along the two partial paths inside the sun ( up to its surface ) and inside the earth ( up to the detector ) .
the corresponding phase differences between the two paths , @xmath52 and @xmath53 @xmath54)$ ] , have been associated to the first path without loss of generality .
the @xmath5 survival probability @xmath40 from eq .
( [ psurv ] ) reads then @xmath55 where the global oscillation phase is given by @xmath56 with the definitions @xmath57 a relevant extension of the @xmath58 formula ( [ pmat ] ) is obtained for @xmath59 oscillations , required to accommodate solar and atmospheric neutrino data @xcite . assuming a third mass eigenstate @xmath60 with @xmath61 , the @xmath59 survival probability can be written as @xmath62 where @xmath63 is the @xmath64 mixing angle , and @xmath65 is given by eq .
( [ pmat ] ) , _ provided that _ the electron density @xmath66 is replaced everywhere by @xmath67 ( see @xcite and refs . therein ) .
such replacement implies that the @xmath59 case is not a simple mapping of the @xmath58 case , and requires specific calculations for any given value of @xmath63 .
we conclude this section by recovering some familiar expressions for @xmath40 , as special cases of eq .
( [ pmat ] ) .
the js limit ( @xmath68 , with complete suppression of oscillations inside matter ) corresponds to @xmath69 and to negligible @xmath70 , @xmath71 .
then , neglecting also @xmath72 , one gets from ( [ pmat ] ) the standard `` vacuum oscillation formula , '' @xmath73 in the msw limit ( @xmath74 ) , the global oscillation phase @xmath75 is very large and @xmath76 on average . furthermore , assuming for @xmath48 a well - known approximation in terms of the `` crossing '' probability @xmath77 between mass eigenstates in matter [ in our notation , @xmath78 , with @xmath79 calculated at the production point ] , one gets from ( [ pmat ] ) and for daytime @xmath80 the so - called parke s formula @xcite , @xmath81
in general , numerical calculations of the @xmath4 transition amplitudes must take into account the detailed @xmath66 profile along the neutrino trajectory , both in the sun and in the earth .
concerning the sun , we take @xmath66 from @xcite ( `` year 2000 '' standard solar model ) . figure [ f1 ] shows such @xmath66 profile as a function of the normalized radius @xmath82 , together with its exponential approximation @xcite @xmath83 , with @xmath84 mol/@xmath85 and @xmath86 . for the exponential density profile
, the neutrino evolution equations can be solved analytically @xcite . in order to calculate the relevant probability @xmath48 and the phase @xmath52 , we have developed several computer programs which evolve numerically the familiar msw neutrino evolution equations @xcite along the sun radius , for generic production points , and for any given value of @xmath31 $ ] ev@xmath2/mev and of @xmath87 .
we estimate a numerical ( fractional ) accuracy of our results better than @xmath88 , as derived by several independent checks . as a first test ,
we integrate numerically the msw equations both in their usual complex form ( 2 real + 2 imaginary components ) and in their bloch form involving three real amplitudes @xcite , obtaining the same results .
we have then repeated the calculations with different integration routines taken from several computer libraries , and found no significant differences among the outputs .
we have optionally considered , besides the exact @xmath66 profile , also the exponential profile , which allows a further comparison of the numerical integration of the msw equations with their analytical solutions , as worked out in @xcite in terms of hypergeometric functions ( that we have implemented in an independent code ) . also in this case , no difference is found between the output of the different codes .
concerning the calculation of the quantities @xmath49 and @xmath53 in the earth , we evolve analytically the msw equations at any given nadir angle @xmath89 , using the technique described in @xcite , which is based on a five - step biquadratic approximation of the density profile from the preliminary reference earth model ( prem ) @xcite and on a first - order perturbative expansion of the neutrino evolution operator .
such analytical technique provides results very close to a full numerical evolution of the neutrino amplitudes , the differences being smaller than those induced by uncertainties in @xmath66 @xcite .
in particular , we have checked that , for @xmath90 ev@xmath2/mev , such differences are @xmath91 . in conclusion , we are confident in the accuracy of our results , which are discussed in the following sections .
figure [ f2 ] shows , in the mass - mixing plane and for standard solar model density , isolines of the difference @xmath92 ( solid curves ) , which becomes zero in the just - so oscillation limit of very small @xmath0 .
the isolines shape reminds the `` lower corner '' of the more familiar msw triangle @xcite .
also shown are isolines of constant resonance radius @xmath93 ( dotted curves ) , defined by the msw resonance condition @xmath94 .
the values of @xmath92 are already sizable ( a few percent ) at @xmath95 , and increase for increasing @xmath0 and for large mixing @xmath96 $ ] , especially in the first octant , where the msw resonance can occur .
the difference between matter effects in the first and in the second octant can lead to observable modifications of the allowed regions in fits to the data @xcite , and to a possible discrimination between the cases @xmath97 and @xmath98 @xcite . in the whole parameter range of fig .
[ f2 ] , it turns out that , within the region of @xmath4 production ( @xmath99 ) , it is @xmath100 ( and thus @xmath101 ) . in the @xmath0 range of fig .
[ f2 ] . ] as a consequence , all the curves of fig .
[ f2 ] do not depend on the specific @xmath4 production point ( as we have also checked numerically ) , and no smearing over the @xmath4 source distribution is needed in the quasi - vacuum regime .
this is a considerable simplification with respect to the msw regime , which involves higher values of @xmath0 and thus shorter ( resonance ) radii , which are sensitive to the detailed @xmath4 source distribution .
figure [ f3 ] shows , in the same coordinates of fig .
[ f2 ] , the isolines of @xmath92 corresponding to the exponential density profile ( dotted curves ) , for which we have used the fully analytical results of @xcite .
( identical results are obtained by numerical integration . ) the solid lines in fig .
[ f3 ] refer to a well - known approximation ( sometimes called `` semianalytical '' ) to such results , which is obtained in the limit @xmath102 at the sun surface ( it is not exactly so for the exponential profile , see fig .
more precisely , the zeroth order expansion of the hypergeometric functions @xcite in terms of the small parameter @xmath103 @xmath104 gives , for @xmath105 , the result @xmath106/ [ \exp(\gamma)-1]$ ] with @xmath107 @xcite , which leads to the qv prescription discussed in @xcite . the differences between the solid and dotted curves in fig .
[ f3 ] are essentially due to the `` solar border approximation '' ( @xmath102 ) assumed in the semianalytical case ; indeed , the differences would practically vanish if the exponential density profile , and thus the `` effective '' sun radius , were unphysically continued for @xmath108 ( not shown )
. such limitations of the semianalytical approximation have been qualitatively suggested by the authors of @xcite but , contrary to their claim , our fig .
[ f3 ] shows explicitly that the semianalytical calculation of @xmath48 represents a reasonable approximation to the analytical one for @xmath12 in _ both _ octants , as also verified in @xcite . a comparison of the results of fig .
[ f2 ] ( true density ) and of fig .
[ f3 ] ( exponential density ) shows that , in the latter case , the correction term @xmath92 tends to be somewhat overestimated , in particular when the semianalytical approximation is used .
we have verified that such bias is dominantly due to the difference ( up to a factor of @xmath109 ) between the true density profile and its exponential approximation around @xmath110 ( see fig .
[ f1 ] ) and , subdominantly , to the details of the density profile shape at the border ( @xmath111 ) . as a consequence , the `` exponential profile '' calculation of @xmath48 ( either semianalytic @xcite or analytic ) tends to shift systematically the onset of solar matter effects to lower values of @xmath0 .
for instance , at @xmath112 ( maximal mixing ) , the value @xmath113 is reached at @xmath114 ev@xmath2/mev for the true density , and at @xmath0 a factor of @xmath109 lower for the exponential density . in order to avoid artificially larger effects at low @xmath0 in neutrino data analyses , one should numerically calculate @xmath48 with the true electron density profile .
the difference between the numerical calculation and the semianalytic approximation is also briefly discussed in @xcite for @xmath115 ev@xmath2/mev ( where @xmath116 ) . concerning the phase factor @xmath70 , we confirm earlier indications @xcite about its smallness , in both cases of true and exponential density . in the latter case ,
the semianalytic approximation gives @xcite , for @xmath117 , the @xmath12-independent result @xmath118-r_\odot\right\}\ , \\\ ] ] where @xmath119 is the euler constant .
numerically , @xmath120 , much smaller than @xmath72 .
as far as @xmath121 ( @xmath122 ) ev@xmath2/mev , we find that @xmath70 differs from such value by less than @xmath123 ( 50% ) , the exact difference depending on the value of @xmath87 ( not shown ) .
therefore , @xmath70 is an order of magnitude smaller than @xmath72 in the whole mass - mixing range considered .
a similar behavior is found by using the true density profile @xmath124 everywhere ) . as for @xmath48
, we find also for @xmath70 no significant dependence on the neutrino production point in the sun core . in conclusion , we find only modest differences between the analytical calculations of @xmath48 , based on the exponential density profile , and its semianalytical approximation .
the difference with respect to the numerical calculation of @xmath48 , based on the standard solar model profile , is instead more pronounced , and leads to a factor of @xmath109 difference in the value of @xmath0 where qv effects start to be significant .
therefore , we recommend the use of the electron profile from the standard solar model , implying numerical calculations of @xmath48 . , calculated for the standard solar model density , upon request from the authors .
] concerning the phase @xmath70 , it simply shifts the global oscillation phase @xmath75 in eq .
( [ pmat ] ) by less than one permill , and thus it can be safely neglected in all current applications .
strong earth matter effects typically emerge in the range where @xmath125 within the mantle ( @xmath126 mol/@xmath85 ) or the core ( @xmath127 mol/@xmath85 ) , as well as in other ranges of mantle - core oscillation interference @xcite , globally corresponding to @xmath128@xmath129 ev@xmath2/mev .
therefore , only marginal effects are expected in the parameter range considered in this work , as confirmed by the results reported in fig .
[ f4 ] . figure [ f4 ] shows isolines of the quantity @xmath130 , which becomes zero in the just - so oscillation limit of very small @xmath0 .
the solid curves corresponds to a nadir angle @xmath131 ( diametral crossing of neutrinos ) and the dotted curves to @xmath132 ( crossing of mantle only ) . for other values of @xmath89 (
not shown ) , the quantity @xmath130 has a comparable magnitude . in the current neutrino jargon ,
the earth effect shown in fig .
[ f4 ] is operative in the lowermost part of the so - called `` low '' msw solution @xcite to the solar neutrino problem , or , from another point of view , to the uppermost part of the vacuum solutions @xcite .
concerning the phase correction @xmath71 ( not shown ) , it is found to be smaller than @xmath133 in the whole mass - mixing plane considered , and thus can be safely neglected . in practical applications ,
the correction term @xmath130 must be time - averaged .
this poses , in principle , a tedious integration problem , since such correction appears , in eq .
( [ pmat ] ) , both in the amplitude of the oscillating term ( @xmath134 ) and in the remaining , non - oscillating term .
while the integration over time can be transformed , for the non - oscillating term , into a more manageable integration over @xmath89 @xcite , this can not be done for the oscillating term , which depends on time both through the prefactor @xmath135 and through the phase @xmath75 ( via eccentricity effects ) . however , in region of the mass - mixing plane where the earth effect is non - negligible ( namely , where @xmath136 few % ) , it turns out that @xmath137 , so that the total amplitude of the oscillating term is small . moreover , as it will be discussed in the next section , energy smearing effects strongly suppress the oscillation factor @xmath138 in the same region .
therefore , for @xmath139 ev@xmath2/mev , the oscillating term in eq .
( [ pmat ] ) is doubly suppressed , and one can safely consider only the non - oscillatory terms .
in the just - so regime , the interference factor @xmath138 in eq .
( [ pmat ] ) can lead to an observable modulation both in the energy and in the time distribution of solar neutrino events .
this modulation gradually disappears as @xmath0 increases , as a consequence of several decoherence effects ( @xmath140 ) @xcite , which are typically dominated by energy smearing @xcite .
the broader the neutrino spectrum , the lower the values of @xmath0 where decoherence becomes important : therefore , it suffices to consider the narrowest spectra ( the so - called be neutrino `` lines '' @xcite ) for illustration purposes .
let us consider a @xmath4 `` line '' energy spectrum @xmath141 , with @xmath142 and @xmath143 .
the energy - averaged oscillating factor , @xmath144 can be written , in the narrow - width approximation @xmath145 , in terms of the fourier transform of the spectrum , @xmath146 more precisely , @xmath147 where @xmath148 , @xmath149 , and @xmath150 . using the @xmath141 profile for the be lines from @xcite ,
we find that the phase correction @xmath151 is smaller than a few permill in the region where the damping factor @xmath152 is greater than a few percent .
if @xmath151 is neglected , the average oscillation factor can be simply written as the oscillation factor at the average energy , times a damping term @xmath152 : @xmath153 a similar approach to smearing effects for neutrino lines was developed in @xcite .
figure [ f5 ] shows the damping factor @xmath152 for the two be neutrino lines at @xmath154 and @xmath155 mev .
the factor is negligible for @xmath156 ev@xmath2/mev , implying that the oscillation pattern is completely smeared out in such range . , where @xmath157 is the line width .
] the onset of smearing effects is shifted to even lower values of @xmath0 for continuous energy spectra . in conclusion ,
the integration over neutrino energy makes the average oscillation factor @xmath158 always negligible in the range where earth matter effects are important , even for the narrow be neutrino lines .
such accidental simplification should not make one forget that , in general , the transition from coherent to incoherent oscillations is a complex phenomenon that deserves further studies ( see , e.g. , @xcite for a recent _ ab initio _ approach ) . finally , we recall that integration over time acts as a further damping factor @xcite . at first order in the eccentricity @xmath159 , one has that @xmath160 $ ] and @xmath161 , where @xmath162 is the orbital phase , and @xmath163 $ ] is the normalized time variable ( @xmath164 at perihelion )
. then the yearly average of @xmath138 [ times the square - law factor @xmath165 can be performed analytically @xcite , giving @xmath166 where @xmath167 is the bessel function , acting as a further damping term for large values of its argument .
notice that the maximum fractional variation of the orbital radius , @xmath168 , is an order of magnitude larger than @xmath169 which , in turn , is larger than the phase corrections @xmath70 and @xmath71 .
therefore , one can safely neglect @xmath72 , @xmath70 and @xmath71 in practical applications involving yearly ( or even seasonal ) averages , as we do in this work . however , for averages over shorter time intervals , such approximation might break down .
in particular , @xmath72 ( @xmath70 ) might be comparable to the monthly ( weekly ) variations of the solar neutrino signal .
the observability of such short - time variations is beyond the present sensitivity of real - time solar @xmath4 experiments and would require , among other things , very high statistics and an extremely stable level of both the signal detection efficiency and of the background .
if such difficult experimental goals will be reached in the future , some of the approximations discussed so far ( and recollected in the next section ) should be revisited and possibly improved .
we have seen in the previous sections that , as @xmath0 increases , the deviations of @xmath48 ( and subsequently of @xmath49 ) from the vacuum value @xmath170 become increasingly important .
we have also seen that the phase corrections @xmath70 and @xmath71 are smaller than @xmath171 , which can in turn be neglected in present applications , so that one can practically take the usual vacuum value for the oscillation phase , @xmath172 .
we think it useful to organize known and less known results through the following approximate expressions for the calculation of @xmath40 , which are accurate to better than 3% with respect to the exact , general formula ( [ pmat ] ) valid at any @xmath0 . for @xmath173 ev@xmath2/mev ,
one can take @xmath174 , and obtain the just - so oscillation formula @xmath175 with @xmath176 . for @xmath177 ev@xmath2/mev ,
one can still take @xmath178 , but since @xmath179 ( quasi - vacuum regime ) one has that @xmath180 where @xmath176 , and @xmath48 has to be calculated numerically ( see also @xcite for earlier versions of the above equation ) .
finally , for @xmath181 ev@xmath2/mev , also earth matter effects are important @xmath182 ; however , this complication is balanced by the disappearance of the oscillating term ( @xmath183 ) due to unavoidable smearing effects , so that the usual msw regime is recovered : @xmath184 time averages are then relatively simple to implement . in eqs .
( [ p1 ] ) and ( [ p2 ] ) , the yearly averages of @xmath138 can be performed analytically [ eq . ( [ faid ] ) ] . in eq .
( [ p3 ] ) , the nighttime average of @xmath185 can be transformed into a more manageable integration over the nadir angle , both for yearly @xcite and for seasonal @xcite averages .
to summarize , the above sequence of equations describes the passage from the regime of just - so to that of msw oscillations , via quasi - vacuum oscillations . in the js regime , oscillations are basically coherent and do not depend on the electron density in the earth or in the sun ( @xmath186 ) . in the msw regime , oscillations are basically incoherent ( @xmath187 ) and , in general , depend on the detailed electron density profile of both the sun and the earth .
in particular , in the msw regime one has to take into account the interplay between the density profile and the neutrino source distribution profile .
the intermediate qv regime is instead characterized by partially coherent oscillations ( with increasing decoherence as @xmath0 increases ) , and by a sensitivity to the electron density of the sun ( but not of the earth ) .
such sensitivity is not as strong as in the msw regime and , in particular , qv effects are independent from the specific @xmath4 production point , which can be effectively taken at the sun center . for the sake of completeness , we mention that , for high values of @xmath3 ( @xmath188 ev@xmath2/mev ) , corresponding to @xmath189 in the sun , the sensitivity to matter effects is eventually lost both in the sun and in the earth ( @xmath174 ) , and one reaches a fourth regime sometimes called of energy - averaged ( ea ) oscillations , which is totally incoherent and @xmath66-independent : @xmath190 such regime , which predicts an energy - independent suppression of the solar neutrino flux , seems to be disfavored ( but perhaps not yet ruled out ) by current experimental data on total neutrino rates . in conclusion , for @xmath3 going from extremely low values to infinity , one can identify four rather different oscillation regimes , @xmath191 each being characterized by specific properties and applicable approximations .
experiments still have to tell us unambiguously which of them truly applies to solar neutrinos .
as discussed in @xcite , in the qv regime the @xmath58 survival probability ( [ p2 ] ) is non - symmetric with respect to the operation @xmath192 , which instead holds for js oscillations [ eq . ( [ p1 ] ) ] .
here we extend such observation to @xmath59 oscillations , under the hypothesis @xmath193 which leads to eq .
( [ p3nu ] ) .
the @xmath58 case is recovered for @xmath194 . equation ( [ p3nu ] ) for @xmath195 preserves the ( a)symmetry properties of @xmath196 under the replacement @xmath192 .
therefore , while @xmath197 [ obtained from eqs .
( [ p3nu ] ) and ( [ p1 ] ) ] is symmetric with respect to the @xmath198 value , the expression of @xmath199 [ obtained from eqs .
( [ p3nu ] ) and ( [ p2 ] ) ] is not .
such properties become evident in the triangular representation of the solar @xmath59 mixing parameter space discussed in @xcite , to which the reader is referred for further details . figure [ f6 ] shows , in the triangular plot , isolines of @xmath199 ( dotted lines ) for @xmath3 close to @xmath200 ev@xmath2/mev , corresponding to about 100 oscillation cycles .
more precisely , the six panels correspond to @xmath201 , with @xmath202 from 0 to @xmath203 in steps of @xmath204 . the dotted isolines are asymmetric with respect to @xmath198 , as is the qv correction @xmath205 . for increasing @xmath206 ( upper part of the triangle )
, the symmetry tends to be restored , mainly because the asymmetric term @xmath207 in eq .
( [ p3nu ] ) is suppressed by the prefactor @xmath208 and , marginally , because the effective electron density @xmath209 in @xmath196 is also suppressed .
is small ( @xmath210 ) , the statement that the quasi - vacuum probability @xmath196 does not depend on the production point in the core ( see sec .
iv ) is not strictly valid , and one should in principle consider also the @xmath4 source profile .
however , since @xmath196 is correspondingly suppressed by @xmath208 , the source smearing effect is numerically small , and one can effectively discard it ( e.g. , by taking the production point at @xmath211 at any @xmath206 ) . ] exact mirror symmetry at any @xmath63 is restored only if the qv corrections are switched off ( solid lines ) ; in such case , @xmath195 is a quadratic form in the coordinates @xmath206 and @xmath212 , and the isolines are conical curves . in any case , @xmath40 becomes typically too low ( too high ) in the central region ( in the corners ) of the triangle where , as a consequence , one should expect only marginal solutions to the solar neutrino deficit .
the asymmetry with respect to @xmath198 also shows up in solar neutrino data fits @xcite .
figure [ f7 ] reports the results of our global ( rates + spectrum + day / night ) three - flavor analysis in the mass - mixing range @xmath213 $ ] ev@xmath2 and @xmath214 $ ] , for several representative values of @xmath215 .
we only show 99% c.l . contours ev@xmath2 . ]
( @xmath216 ) for the sake of clarity .
the theoretical @xcite and experimental @xcite inputs , as well as the @xmath217 statistical analysis @xcite , are the same as in ref .
@xcite ( where msw solutions were studied ) . here
, however , the range of @xmath11 is lower , in order to show the smooth transition from the msw solutions to the qv and finally the js ones as @xmath11 is decreased . in particular , the solutions shown in fig . [ f7 ] represent the continuation , at low @xmath218 , of the low msw solutions shown in fig .
10 of @xcite ( panel by panel ) . in the plane of fig .
[ f7 ] ( @xmath219 ) is reached at @xmath220 . for @xmath221 ev@xmath2 ( msw regime ) ,
the minimum value is @xmath222 @xcite for the same input data . in order to match the results of fig .
[ f7 ] in this work and of fig . 10 in @xcite
, we have adopted the _
absolute _ minimum ( @xmath222 ) to draw the @xmath223 contours in fig .
[ f7 ] . ] as anticipated in the comments to fig .
[ f6 ] , the mirror asymmetry around @xmath224 decreases for decreasing @xmath11 ( js regime ) ; a little asymmetry is still present even at @xmath225 ev@xmath2 , where the gallium rates ( sensitive to @xmath10 as low as @xmath226 mev ) start to feel qv effects . in the region where qv effects are important
, the solutions are typically shifted in the second octant @xmath227 ) , since the gallium rate is suppressed too much in the first octant ( see also @xcite ) .
a similar drift was found for the low msw solution @xcite . at any @xmath218
, the asymmetry decreases at large values of @xmath63 ( @xmath228 ) which , however , are excluded by the combination of accelerator and reactor data @xcite , unless the second mass square difference @xmath229 turns out to be in the lower part of the sensitivity range of the chooz experiment @xcite ( @xmath230 ev@xmath2 ) . for @xmath194 ,
the standard two - flavor case is recovered , and the results are comparable to those found in @xcite .
the super - kamiokande spectrum plays only a marginal role in generating the mirror asymmetry of the solutions in fig .
[ f7 ] , since the modulation of qv effects in the energy domain is much weaker than the one generated by the oscillation phase @xmath75 .
we find that , at any given @xmath231 ev@xmath2 , the @xmath217 difference at symmetric @xmath12 values is less than @xmath232 for the 18-bin spectrum data fit .
therefore , qv effects are mainly probed by total neutrino rates at present .
we think it is not particularly useful to discuss more detailed features of the current qv solutions , such as combinations of spectral data or rates only , fits with variations of _ hep _ neutrino flux , etc .
( which were instead given in @xcite for msw solutions ) .
in fact , while the shapes of current msw solutions are rather well - defined , those of js or qv solutions are still very sensitive to small changes in the theoretical or experimental input .
therefore , a detailed analysis of the `` fine structure '' of the qv solutions in fig .
[ f7 ] seems unwarranted at present . finally , in fig .
[ f8 ] we show sections of the allowed @xmath59 solutions ( at 99% c.l . ) in the triangle representation , for six selected ( increasing ) values of @xmath11 .
solutions are absent or shrunk at @xmath233 , where the theoretical @xmath4 flux underestimates the gallium and water - cherenkov data .
the lowest value of @xmath11 ( @xmath234 ev@xmath2 ) falls in the js regime , so that the ring - like allowed region ( which resembles the curves of iso-@xmath40 in fig .
[ f6 ] ) is symmetric with respect to the vertical axis at @xmath235 .
however , as @xmath11 increases and qv effects become operative , the solutions become more and more asymmetric , and shifted towards the second octant of @xmath12 @xcite .
figures [ f7 ] and [ f8 ] in this work , as well as fig . 10 in @xcite , show that solar neutrino data , by themselves , put only a weak upper bound on the mixing angle @xmath63 .
much tighter constraints are set by reactor data @xcite , unless the second mass square difference @xmath229 happens to be @xmath236 ev@xmath2 ( which seems an unlikely possibility ) . in any case , qv effects are operative also for a small ( or zero ) value of @xmath206 .
we have presented a thorough analysis of solar neutrino oscillations in the `` quasi - vacuum '' oscillation regime , intermediate between the familiar just - so and msw regimes .
the qv regime is increasingly affected by matter effects for increasing values of @xmath11 .
we have calculated such effects both in the sun and in the earth , and discussed the accuracy of various possible approximations .
we have implemented the qv oscillation probability in a full three - flavor analysis of solar neutrino data , obtaining solutions which smoothly join ( at @xmath237 ev@xmath2 ) the low msw regions found in @xcite for the same input data .
the asymmetry of qv effects makes such solutions different for @xmath97 and @xmath98 , the two cases being symmetrized only in the just - so oscillation limit of small @xmath11 .
y. suzuki for the super - kamiokande collaboration , in _ lepton - photon 99 _ , proceedings of the xix international symposium on photon and lepton interactions at high energies ( stanford , california , u.s.a . , 1999 ) , to appear ; transparencies available at http://lp99.slac.stanford.edu fig .
[ f1 ] . radial profile of the electron density in the sun from the standard solar model ( solid line ) , together with its exponential approximation ( dashed line ) .
isolines of the solar correction term @xmath92 in the mass - mixing plane ( solid curves ) , for standard solar model density @xcite .
isolines of msw resonance radii ( dashed curves ) are also shown .
[ f3 ] . as in fig .
[ f2 ] , but for the exponential density profile .
dotted and solid lines correspond to analytical calculations and to their semianalytical approximation , respectively .
isolines of the earth correction term @xmath130 in the mass - mixing plane , for prem @xcite density .
the solid and dotted lines correspond to a nadir angle equal to @xmath239 and @xmath240 , respectively .
[ f5 ] . damping factor for the oscillating term , due to the finite
be line width .
see the text for details .
three - flavor oscillations : isolines of @xmath40 in the triangular representation of the solar @xmath4 mixing parameter space , for representative values of the oscillation phase @xmath75 .
the dotted and solid lines refer , respectively , to calculations with and without quasi - vacuum effects .
when such effects are included , the mirror symmetry @xmath192 is broken .
three - flavor oscillations : contours at 99% c.l .
, as derived from a global analysis of neutrino data for @xmath33 ev@xmath2 ( quasi - vacuum regime ) .
the asymmetry of the solutions is evident for increasing values of @xmath11 .
the solutions represent the continuation ( at low @xmath11 ) of the msw 99% c.l .
regions reported in fig .
10 of @xcite .
sections of the @xmath59 allowed volume ( 99% c.l . ) for six representative values of @xmath11 ( 0.66 , 2.36 , 6.42 , 8.76 , 10.8 , and @xmath241 ev@xmath2 ) , shown in the solar triangle plot . | we discuss in detail solar neutrino oscillations with @xmath0 in the range @xmath1 $ ] ev@xmath2/mev . in this range , which interpolates smoothly between
the so - called `` just - so '' and `` mikheyev - smirnov - wolfenstein '' oscillation regimes , neutrino flavor transitions are increasingly affected by matter effects as @xmath3 increases . as a consequence ,
the usual vacuum approximation has to be improved through the matter - induced corrections , leading to a `` quasi - vacuum '' oscillation regime .
we perform accurate numerical calculations of such corrections , using both the true solar density profile and its exponential approximation .
matter effects are shown to be somewhat overestimated in the latter case .
we also discuss the role of earth crossing and of energy smearing .
prescriptions are given to implement the leading corrections in the quasi - vacuum oscillation range .
finally , the results are applied to a global analysis of solar @xmath4 data in a three - flavor framework . |
the cepheus a region is an active site of galactic star - formation , which contains one of the very few well - documented examples ( hw 2 ) of a high - mass protostellar system where a disk - like flattened structure has been detected ( patel et al .
the distance to cepheus a has traditionally been very uncertain with estimates ranging from 300 pc ( migenes et al .
1992 ) to 900 pc ( moreno - corral et al . 1993 ) .
recently , moscadelli et al . ( 2009 ) obtained a direct parallax measurement based on multi - epoch observations of a methanol maser feature associated with hw 2 , using the vlba telescope . the corresponding result ( @xmath3 = 700 @xmath1 40 pc ) significantly reduced the uncertainty on the distance to cepheus a , but was based on a single set of observations of a single source . in the present paper ,
the sixth of our series dedicated to very long baseline array determinations of distances to nearby star - forming regions ( see loinard et al . 2005 , 2007 , 2008 ; torres et al . 2007 , 2009 ;
dzib et al .
2010 for the previous papers of the series ) , we will provide an entirely independent measurement of the distance to cepheus a based on multi - epoch observations of the continuum emission associated with the radio source hw 9 .
hw 9 was first reported by hughes ( 1991 ) as the ninth radio source in cepheus a ( following previous detections in the same region reported in hughes & wouterloot 1984 ) .
hughes ( 1991 ) found that the source was very compact as well as highly variable , and interpreted its radio emission in terms of gyrosynchrotron radiation ( see also hughes et al .
1995 and garay et al .
1996 ) . in high angular resolution maps ( figure [ fig : pm ] ) ,
hw 9 appears as an isolated , featureless source located about 5@xmath4 to the south - east of the well - studied high - mass object hw 2 .
the visual extinction toward hw 9 is at least 23 magnitudes ( pravdo et al .
2009 ) and could be as high as 100 magnitudes ( hughes 1991 ; pravdo et al .
2009 ) . as a consequence
, no infrared or visual counterpart has ever been reported for hw 9 , no spectral classification is available , and the very nature of the source remains very unclear . because of its possible association with an h ii region , hughes ( 1991 ) proposed that hw 9 might be a b3 star .
garay et al . ( 1996 ) , however , argued that hw 9 is more likely to be a low mass stellar object because of its radio flaring activity .
the observations reported in this paper will provide some additional constraints on the nature of this enigmatic source .
in total , ten continuum observations collected at a wavelength of 3.6 cm ( @xmath5 = 8.42 ghz ) will be reported here . the first one ( obtained in february 2007 )
was designed as a detection experiment . following the successful detection of the source in that first observation
, we initiated a series of nine observations starting in october 2007 .
the separation between successive observations in this subsequent data set was about three months , so the last observation occurred in november 2009 .
our main phase calibrator for all epochs was j2302 + 6405 , located at an angular distance of 2.19 degrees from the target . to improve the quality of phase calibration
, we also observed secondary phase calibrators .
for the first epoch we used the quasars j2258 + 5719 , j2223 + 6249 , and j2322 + 6911 ( at angular distances of 4.72 , 3.90 and 7.63 degrees , respectively , from the target ) .
for the last nine observations , j2322 + 6911 was replaced by the somewhat more nearby quasar j2309 + 6820 at 6.45 degrees from the target ( see figure [ fig : calibrators ] for the relative positions of the calibrators ) . the faint quasar j2254 + 6209 , located at 0.26 degrees from hw 9
was also observed during the last nine observations .
it could not be used as a calibrator because it is faint , resolved , and variable , but provided a very useful check on the overall quality of the astrometry .
each observation consisted of series of cycles with two minutes spent on source , and one minute spent on the main phase calibrator j2302 + 6405 .
roughly every 30 minutes , we observed the secondary calibrators , spending one minute on each .
in addition , geodetic blocks consisting of observations of about two dozen calibrators spread over the entire visible sky were collected at the beginning , the middle , and the end of each of our multi - epoch observations . with these overheads ,
five and three hours were spent on source during the first observation , and each of the subsequent epochs , respectively .
the data were edited and calibrated using the astronomical image processing system ( aips ; greisen 2003 ) .
the basic data reduction followed the standard vlba procedure for phase - referenced observations , including the multi - calibrator schemes and tropospheric and clock corrections .
these calibrations were described in detail by loinard et al .
( 2007 ) , torres et al .
( 2007 ) and in dzib et al .
after their calibration , the visibilities were imaged with a pixel size of 50 @xmath6as using a weighting scheme intermediate between natural and uniform ( robust = 0 in aips ) . the r.m.s .
noise levels in the final images were 0.08 0.12 mjy beam@xmath7 . from these images ,
the source position was determined using a two - dimensional fitting procedure ( task jmfit in aips ) .
hw 9 was detected in four of the ten observed epochs .
these detections imply brightness temperatures reaching 10@xmath8 k , and clearly demonstrate that the radio emission is of non - thermal ( presumably gyrosynchrotron ) origin .
the source is very variable , reaching a maximum flux of about 2.8 mjy in the first and eighth epochs ( figure 3 ) while remaining undetectable at levels below @xmath90.22 mjy in several observations .
this corresponds to a maximum - to - minimum flux ratio in excess of at least 14 .
the combination of the high brightness and variability of the radio emission indicates that hw 9 is a flaring star with coronal emission , as previously suggested ( e.g. , garay et al .
1996 ) .
the positions of hw 9 measured from our vlba observations were modeled as a combination of a trigonometric parallax ( @xmath0 ) and proper motion ( @xmath6 assumed to be uniform ) following loinard et al .
the barycentric coordinates of the earth as well as the julian date appropriate for each observation were calculated using the multi - year interactive computer almanac ( mica ) distributed as a cd rom by the us naval observatory .
the reference epoch was taken at jd 2454590.95 @xmath10 j2008.36 , the mean epoch of our our detections .
the best fit to the data ( figure 4 ) yields the following astrometric elements : @xmath11 the post - fit r.m.s .
is 0.04 mas in declination , indicating that no systematic errors remain along that axis . in right ascension , however , a systematic contribution of 0.15 mas has to be added quadratically to the errors given by jmfit to yield a reduced @xmath12 of 1 .
these systematic errors might be related to the east - west structure present in the images of the main phase calibrator , and have been included in all the uncertainties quoted in the present paper .
[ fig : flux ]
the parallax found here for hw 9 is in excellent agreement with the value found by moscadelli et al .
( 2009 ) for a methanol maser associated with the source hw 2 in cepheus a ( @xmath0 = 1.43 @xmath1 0.08 mas ) .
this concordance between two independent results confirms the reliability of vlba ( and , more generally , of very long baseline interferometry ) measurements of trigonometric parallaxes . by combining the two results , we constrain the parallax of the cepheus a region to be 1.43 @xmath1 0.06 , corresponding to a distance @xmath3 = 700@xmath13 pc . the astrometry carried out by moscadelli et al .
( 2009 ) yielded accurate proper motions for a methanol maser associated with the massive young stellar object hw 2 .
the velocity of such masers are usually believed to agree with that of their associated young stars to better than 3 km s@xmath7 ( moscadelli et al .
2002 , 2009 ) .
the best proper motion obtained by moscadelli et al .
( 2009 ) for the methanol maser in the hw 2 region was @xmath14 mas yr@xmath7 and @xmath15 mas yr@xmath7 . by comparing these figures with our own results ( section 3 )
, it can be seen that the proper motions of hw 2 and hw 9 are consistent within @xmath16 in right ascension , but differ by more than @xmath17 in declination ( see figure [ fig : pm ] ) .
the difference ( @xmath18 ) corresponds to a velocity difference @xmath19 = 6.2 @xmath1 0.7 km s@xmath7 , which likely reflects the combination of ( i ) a @xmath9 3 km s@xmath7 difference between the velocity of hw 2 and that of its associated maser , and ( ii ) a few km s@xmath7 difference between the space velocities of hw 2 and hw 9 due to the expected velocity dispersion within the cepheus a region . using vera ( vlbi exploration of radio astrometry ) observations of water masers , hirota et al .
( 2008 ) estimated the distance ( @xmath3 = 764 @xmath1 27 pc ) to the young massive stellar object iras@xmath20 . in projection , this source is located at 4.4 degrees ( about 55 pc ) from cepheus a and both regions are part of the cepheus cassiopeia molecular cloud complex .
the difference in distance appears to be similar to the separation on the plane of the sky , suggesting that the cepheus
cassiopeia is about as deep as it is wide .
we note that similar depths have been found for other molecular cloud complexes ( i.e. straiys et al .
2003 ) as mentioned in section 1 , the exact nature of hw 9 remains unclear .
the present detection with the vlba clearly demonstrates that the radio emission is of non - thermal origin , and traces a flaring corona .
this is in agreement with previous analyses of vla observations ( e.g. garay et al . 1996 ) , and with the interpretation proposed by pravdo et al . ( 2009 ) of the x - ray emission toward hw 9 .
high - mass stars are not expected to power active coronas because they are fully radiative . as a consequence
, the dynamo mechanism can not operate in their interior and they do not generate the strong magnetic fields required to maintain a corona .
our observations , therefore , indicate that hw 9 is either a low mass ( t tauri ) or an intermediate mass ( herbig ae / be ) young star . for t tauri stars ,
the x - ray and radio emissions appear to be well correlated ( e.g. benz & gdel 1994 ) .
the typical x - ray to radio luminosity ratio for these objects is @xmath21 @xmath9 10@xmath22 hz , with a dispersion around the relation of about one dex ( gdel 2004 ; benz & gdel 1994 ) . for herbig ae
/ be stars , on the other hand , hamidouche et al .
( 2008 ) found a relation @xmath23 hz , but with significant dispersion .
in particular , values of @xmath24 as high as @xmath25 hz were found . using radio observations with lower angular resolution
than those presented here , pravdo et al . ( 2009 ) found @xmath26 hz for hw 9 . combining the data presented here with the x - ray observations of pravdo et al .
( 2009 ) , we find a similar @xmath24 ratio of @xmath27 hz .
this places hw 9 near the lower - end of the @xmath24 relation for t tauri stars , and near the upper - end of the relation for herbig ae / be stars .
our observations , therefore , do not strongly constrain the mass of hw 9 beyond the fact that it is not a massive object .
the cepheus a region is a site of very recent star - formation ; hw 2 , in particular , is believed to be a very young stellar object .
the source hw 9 studied here is located only 5@xmath4 ( less than 0.02 pc ) away , and is very likely to be coeval with hw 2 .
the very high extinction toward hw 9 mentioned in section 1 further reinforces the idea that it is a highly embedded , very young object . indeed , pravdo et al .
( 2009 ) suggested that hw 9 might be a class 0/i proto - stellar object .
evidence for magnetic activity around such young objects has been reported in a limited number of cases ( see gdel 2002 , for a discussion ) .
we note , however , that coronal radio emission has never been detected from class 0 sources , and has been conclusively established for only a few class i objects ( forbrich et al .
2009 ; dzib et al .
the likely reason for this paucity is related to the existence around very young objects of partially ionized winds which generate optically thick free - free radio emission surrounding the young stellar source .
any non - thermal coronal radio emission from the young star itself would be absorbed in the optically thick layers of the winds , and would never reach the observer .
the amount of free - free emission associated with hw 9 , however , appears to be very limited . in particular , vla observations at 6 cm
have sometimes failed to detect emission from hw 9 at levels of about 0.15 mjy ( hughes & wouterloot 1984 ; garay et al .
thus , if hw 9 is indeed a very young object , it is one with very limited ejection activity , and this lack of strong winds would help explain the presence of the coronal emission detected here .
in this paper , we reported on vlba observations of the young stellar object hw 9 in the cepheus a star - forming region .
these data have been used to provide an independent confirmation that cepheus a is located at a distance of 700 pc , and to reduce the uncertainty on that distance from about 40 pc down to about 30 pc .
while it is clear that hw 9 is young and less massive than @xmath9 6 @xmath28 , its exact nature remains unclear .
in particular , its radio properties are intermediate between those of low - mass t tauri stars and those of intermediate mass herbig ae / be objects .
is grateful to the guggenheim foundation for financial support .
, l.l . and l.f.r .
acknowledge the financial support of dgapa , unam and conacyt , mxico , while r.m.t .
acknowledges support by the deutsche forschungsgemeinschaft ( dfg ) through the emmy noether research grant vl 61/3 - 1 .
the national radio astronomy observatory is a facility of the national science foundation operated under cooperative agreement by associated universities , inc . | using the very long baseline array ( vlba ) , we have observed the radio continuum emission from the young stellar object hw 9 in the cepheus a star - forming region at ten epochs between 2007 february and 2009 november . due to its strong radio variability ,
the source was detected at only four of the ten epochs . from these observations ,
the trigonometric parallax of hw 9 was determined to be @xmath0 = 1.43 @xmath1 0.07 mas , in excellent agreement with a recent independent vlba determination of the trigonometric parallax of a methanol maser associated with the nearby young stellar source hw 2 ( @xmath0 = 1.43 @xmath1 0.08 mas ) .
this concordance in results , obtained in one case from continuum and in the other from line observations , confirms the reliability of very long baseline array trigonometric parallax measurements . by combining the two results ,
we constrain the distance to cepheus a to be 700@xmath2 pc , an uncertainty of 3.5% . |
details of the star experiment can be found in @xcite . for this analysis ,
events with primary vertex within @xmath5 cm of the tpc center are used .
the au+au events are divided into 7 centrality classes as in @xcite .
a large @xmath1 trigger particle is selected within the pseudo - rapidity and transverse momentum ranges of @[email protected] and 4@xmath0@xmath3@xmath06 gev/@xmath2 .
charged hadrons within 0.15@xmath0@xmath1@xmath04 gev/@xmath2 and @[email protected] are associated with the trigger particle in @xmath8 and @xmath9 . acceptance and tracking efficiency are corrected for the associated particles .
background is constructed by an event - mixing technique .
the elliptic flow contribution , @xmath10 , is included .
residual background effect is minimized by averaging the lowest data points ( mostly within 0.9@xmath0@[email protected] ) to zero .
figure [ fig1 ] shows the background subtracted @xmath12 distributions .
jet - like structures are observed in p+p and on the near side in au+au collisions .
we define near side as within @[email protected] and @[email protected] and away side as within @[email protected] and @[email protected] .
we integrate the correlation signals as a measure of the jet charged hadron multiplicities .
the near side signal is corrected for two - particle @xmath15 acceptance .
we deduce the total scalar @xmath1 by applying a multiplicative factor of 1.5 to the associated particles to account for the neutral particles ; we include the trigger particle @xmath3 for the near jet .
systematic uncertainties on our measurements come from three major sources : those in ( i ) elliptic flow corrections , ( ii ) background normalization , and ( iii ) efficiency corrections .
the estimated systematic uncertainties are shown on the results where indicated .
figure [ fig2 ] shows the charged hadron multiplicity and total scalar @xmath1 on both the near and away side in p+p and as a function of centrality in au+au collisions .
the trigger particle is included in the near side results .
we note that the away side measurements are those within the acceptance only , whereas those measured on the near side correspond to a nearly complete jet . as seen from fig .
[ fig2 ] , by triggering on a large @xmath1 particle , higher energy jets are selected in central au+au than in p+p collisions .
the charged hadron multiplicity and total scalar @xmath1 are found to increase from p+p to central au+au collisions for both the near and away side .
figure [ fig3](a - b ) shows the @xmath1 distributions of associated particles in p+p , peripheral 80 - 40% and central 5% au+au collisions .
while general agreement is found between p+p and peripheral au+au , the central au+au results differ from p+p , most significantly for the away side .
the away side in central au+au is depleted at large @xmath1 and enhanced at low @xmath1 compared to p+p collisions
. the ratios of au+au to p+p distributions are depicted in fig .
[ fig3](c - d ) .
it appears that jets are modified in the medium created in au+au collisions ; the modification results in more associated particles on the near side and moves energy from high to low @xmath1 on the away side .
figure [ fig4 ] shows the centrality dependence of @xmath16 of associated charged hadrons on the away side .
also shown is the @xmath16 of inclusive hadrons that make up the medium .
the away side @xmath16 is significantly larger than that for inclusive hadrons in p+p and peripheral au+au collisions , and drops significantly with centrality in au+au collisions , approaching the medium hadron @xmath16 .
this may indicate a progressive equilibration between the associated hadrons and the medium from peripheral to central collisions .
borghini _ et al .
_ estimated , on a statistical basis , effect on two - particle correlation due to global momentum conservation @xcite .
such an effect would yield a correlation of @xmath17 , where @xmath18 is the near side jet @xmath1 and the summations run over all particles ( denominator ) and those in the acceptance only ( numerator ) .
an estimate of such a correlation with a constant offset is shown as the band in fig .
[ fig1 ] ( see also ref .
the curve shows a free fit to the data with the sum of a gaussian ( for near side ) , a @xmath19 term ( for away side ) , and a constant offset ; it appears to describe the away side distributions well , especially for central collisions .
this may suggest that the away side associated particles are reaching equilibration with the medium , in consistency with the softening of the @xmath1 spectra shown in fig .
[ fig3](b ) .
statistical reconstruction of charged hadron jets is performed for p+p and au+au collisions at rhic by correlating charged hadrons in the @xmath1 range of 0.15@xmath0@xmath1@xmath04 gev/@xmath2 to large @xmath1 particles within 4@xmath0@xmath3@xmath06 gev/@xmath2 and subtracting mixed - event background .
preliminary results on charged hadron multiplicity , total scalar @xmath1 , and associated charged hadron @xmath1 distribution are presented for both the near and away side .
it is found , with the same @xmath3 leading particle in the final state , that the total scalar @xmath1 is larger in central au+au than in p+p collisions , and increases gradually with centrality .
the @xmath1 distribution of associated charged hadrons on the away side is found to be significantly softened in central au+au with respect to p+p collisions , suggesting a significant equilibration of those hadrons with the medium .
the results are consistent with modification of jets in nuclear medium created at rhic .
et al . _ ( phenix ) , phys .
lett . * 88 * ( 2002 ) 022301 ; c. adler , _ et al . _ ( star ) , phys . rev . lett .
* 89 * ( 2002 ) 202301 ; s.s .
et al . _ ( phenix ) , phys .
* 91 * ( 2003 ) 072301 ; j. adams , _ et al . _
( star ) , phys .
* 91 * ( 2003 ) 172302 . c. adler ,
_ et al . _ ( star ) , phys .
lett . * 90 * ( 2003 ) 082302 .
b.b . back , _
( phobos ) , phys .
* 91 * ( 2003 ) 072302 ; s.s .
adler , _ et al . _
( phenix ) , phys .
* 91 * ( 2003 ) 072303 ; j. adams , _ et al . _
( star ) , phys .
* 91 * ( 2003 ) 072304 ; i. arsene , _ et al . _
( brahms ) , phys .
* 91 * ( 2003 ) 072305 .
m. gyulassy and m. plmer , phys .
b * 243 * ( 1990 ) 432 ; x .-
n . wang and m. gyulassy , phys .
* 68 * ( 1992 ) 1480 ; m. gyulassy and x .-
wang , nucl . phys .
b * 420 * ( 1994 ) 583 ; x .-
wang , m. gyulassy , and m. plmer , phys .
d * 51 * ( 1995 ) 3436 ; r. baier , d. schiff , and b.g .
zakharov , ann .
nucl . part .
* 50 * ( 2000 ) 37 . | charged hadrons ( 0.15@xmath0@xmath1@xmath04 gev/@xmath2 ) associated with a large @xmath1 trigger particle ( 4@xmath0@xmath3@xmath06 gev/@xmath2 ) are statistically reconstructed in the large acceptance star tpc for p+p and au+au collisions at @xmath4=200 gev .
preliminary results on transverse momentum distributions of these hadrons are presented , separately on the near and away side from the trigger particle .
an increase in the multiplicity and total scalar @xmath1 of these hadrons from p+p to central au+au collisions is observed .
the away side distribution is found to be significantly softened in central au+au with respect to p+p collisions .
the results are consistent with modification of jets in heavy ion collisions at rhic . at the relativistic heavy ion collider ( rhic ) ,
yield of large transverse momentum ( @xmath1 ) hadrons per binary nucleon - nucleon collision was found to be suppressed in central au+au relative to p+p collisions @xcite ; the strong jet - like back - to - back correlation between large @xmath1 hadrons , as observed in p+p , was found not present in central au+au collisions @xcite ; d+au collisions were found to be similar to p+p collisions @xcite .
these results are consistent with the picture where hard - scattered partons are similarly produced in au+au and p+p but subsequently lose energy in central au+au collisions due to final state interactions @xcite .
reconstruction of the lost energy ( or jets of hadrons ) @xcite will therefore serve as a strong experimental confirmation of the energy loss picture .
moreover , by studying the amount of energy loss and how the energy is distributed , one may learn experimentally about the medium density , the underlying energy loss mechanism(s ) , and the degree of equilibration between the energy and the medium .
unlike in elementary collisions , jet reconstruction in central heavy ion collisions poses an experimental challenge due to the overwhelming background multiplicity at low @xmath1 in these collisions .
we report our attempt to statistically reconstruct jets of charged hadrons by subtracting the background in the large acceptance star tpc , and present preliminary results . |
in many applications one fits a parametrized curve described by an implicit equation @xmath0 to experimental data @xmath1 , @xmath2 . here
@xmath3 denotes the vector of unknown parameters to be estimated .
typically , @xmath4 is a polynomial in @xmath5 and @xmath6 , and its coefficients are unknown parameters ( or functions of unknown parameters ) .
for example , a number of recent publications @xcite are devoted to the problem of fitting quadrics @xmath7 , in which case @xmath8 is the parameter vector .
the problem of fitting circles , given by equation @xmath9 with three parameters @xmath10 , also attracted attention @xcite .
we consider here the problem of fitting general curves given by implicit equations @xmath0 with @xmath11 being the parameter vector .
our goal is to investigate statistical properties of various fitting algorithms .
we are interested in their biasedness , covariance matrices , and the cramer - rao lower bound .
first , we specify our model .
we denote by @xmath12 the true value of @xmath3 .
let @xmath13 , @xmath2 , be some points lying on the true curve @xmath14 .
experimentally observed data points @xmath15 , @xmath2 , are perceived as random perturbations of the true points @xmath13 .
we use notation @xmath16 and @xmath17 , for brevity .
the random vectors @xmath18 are assumed to be independent and have zero mean .
two specific assumptions on their probability distribution can be made , see @xcite : * _ cartesian model _ :
each @xmath19 is a two - dimensional normal vector with covariance matrix @xmath20 , where @xmath21 is the identity matrix . * _ radial model _ :
@xmath22 where @xmath23 is a normal random variable @xmath24 , and @xmath25 is a unit normal vector to the curve @xmath14 at the point @xmath26 .
our analysis covers both models , cartesian and radial . for simplicity , we assume that @xmath27 for all @xmath28 , but note that our results can be easily generalized to arbitrary @xmath29
. concerning the true points @xmath30 , @xmath2 , two assumptions are possible .
many researchers @xcite consider them as fixed , but unknown , points on the true curve . in this case
their coordinates @xmath13 can be treated as additional parameters of the model ( nuisance parameters ) .
chan @xcite and others @xcite call this assumption a _
functional model_. alternatively , one can assume that the true points @xmath30 are sampled from the curve @xmath31 according to some probability distribution on it .
this assumption is referred to as a _ structural model _ @xcite .
we only consider the functional model here .
it is easy to verify that maximum likelihood estimation of the parameter @xmath3 for the functional model is given by the orthogonal least squares fit ( olsf ) , which is based on minimization of the function _ 1 ( ) = _ i=1^n [ d_i()]^2 [ fmain1 ] where @xmath32 denotes the distance from the point @xmath26 to the curve @xmath0 .
the olsf is the method of choice in practice , especially when one fits simple curves such as lines and circles .
however , for more general curves the olsf becomes intractable , because the precise distance @xmath33 is hard to compute .
for example , when @xmath4 is a generic quadric ( ellipse or hyperbola ) , the computation of @xmath33 is equivalent to solving a polynomial equation of degree four , and its direct solution is known to be numerically unstable , see @xcite for more detail .
then one resorts to various approximations .
it is often convenient to minimize _ 2 ( ) = _ i=1^n [ p(x_i , y_i;)]^2 [ fmain2 ] instead of ( [ fmain1 ] ) .
this method is referred to as a ( simple ) _ algebraic fit _ ( af ) , in this case one calls @xmath34 the _ algebraic distance _
@xcite from the point @xmath1 to the curve .
the af is computationally cheaper than the olsf , but its accuracy is often unacceptable , see below .
the simple af ( [ fmain2 ] ) can be generalized to a _ weighted algebraic fit _ , which is based on minimization of _ 3 ( ) = _
i=1^n w_i [ p(x_i , y_i;)]^2 [ fmain3 ] where @xmath35 are some weights , which may balance ( [ fmain2 ] ) and improve its performance .
one way to define weights @xmath36 results from a linear approximation to @xmath33 : @xmath37 where @xmath38 is the gradient vector , see @xcite .
then one minimizes the function _ 4 ( ) = _ i=1^n [ fmain4 ] this method is called the _ gradient weighted algebraic fit _ ( graf ) .
it is a particular case of ( [ fmain3 ] ) with @xmath39 .
the graf is known since at least 1974 @xcite and recently became standard for polynomial curve fitting @xcite .
the computational cost of graf depends on the function @xmath40 , but , generally , the graf is much faster than the olsf .
it is also known from practice that the accuracy of graf is almost as good as that of the olsf , and our analysis below confirms this fact . the graf is often claimed to be a _ statistically optimal _ weighted algebraic fit , and we will prove this fact as well . not much has been published on statistical properties of the olsf and algebraic fits , apart from the simplest case of fitting lines and hyperplanes @xcite .
chan @xcite , berman and culpin @xcite investigated circle fitting by the olsf and the simple algebraic fit ( [ fmain2 ] ) assuming the structural model .
kanatani @xcite used the cartesian functional model and considered a general curve fitting problem .
he established an analogue of the rao - cramer lower bound for unbiased estimates of @xmath3 , which we call here kanatani - cramer - rao ( kcr ) lower bound .
he also showed that the covariance matrices of the olsf and the graf attain , to the leading order in @xmath41 , his lower bound .
we note , however , that in most cases the olsf and algebraic fits are _ biased _
@xcite , hence the kcr lower bound , as it is derived in @xcite , does not immediately apply to these methods . in this paper
we extend the kcr lower bound to biased estimates , which include the olsf and all weighted algebraic fits .
we prove the kcr bound for estimates satisfying the following mild assumption : * precision assumption*. for precise observations ( when @xmath42 for all @xmath43 ) , the estimate @xmath44 is precise , i.e. ( |*x*_1 , , |*x*_n ) = | [ tass ] it is easy to check that the olsf and algebraic fits ( [ fmain3 ] ) satisfy this assumption .
we will also show that all unbiased estimates of @xmath44 satisfy ( [ tass ] ) .
we then prove that the graf is , indeed , a statistically efficient fit , in the sense that its covariance matrix attains , to the leading order in @xmath41 , the kcr lower bound . on the other hand ,
rather surprisingly , we find that graf is not the only statistically efficient algebraic fit , and we describe all statistically efficient algebraic fits .
finally , we show that kanatani s theory and our extension to it remain valid for the radial functional model .
our conclusions are illustrated by numerical experiments on circle fitting algorithms .
recall that we have adopted the functional model , in which the true points @xmath30 , @xmath43 , are fixed .
this automatically makes the sample size @xmath45 fixed , hence , many classical concepts of statistics , such as consistency and asymptotic efficiency ( which require taking the limit @xmath46 ) lose their meaning .
it is customary , in the studies of the functional model of the curve fitting problem , to take the limit @xmath47 instead of @xmath46 , cf.@xcite .
this is , by the way , not unreasonable from the practical point of view : in many experiments , @xmath45 is rather small and can not be ( easily ) increased , so the limit @xmath48 is of little interest . on the other hand , when the accuracy of experimental observations is high ( thus , @xmath41 is small ) , the limit @xmath49 is quite appropriate .
now , let @xmath50 be an arbitrary estimate of @xmath3 satisfying the precision assumption ( [ tass ] ) . in our analysis we will always assume that all the underlying functions are regular ( continuous , have finite derivatives , etc . ) , which is a standard assumption @xcite .
the mean value of the estimate @xmath44 is e ( ) = ( * x*_1, ,*x*_n ) _
i=1^n f(*x*_i ) d*x*_1d*x*_n [ et ] where @xmath51 is the probability density function for the random point @xmath26 , as specified by a particular model ( cartesian or radial ) .
we now expand the estimate @xmath52 into a taylor series about the true point @xmath53 remembering ( [ tass ] ) : ( * x*_1 , , * x*_n ) = | + _ i=1^n
_ i ( * x*_i - |*x*_i ) + o(^2 ) [ texpand ] where _
i = _ * x*_i ( |*x*_1 , , for the gradient with respect to the variables @xmath54 .
in other words , @xmath55 is a @xmath56 matrix of partial derivatives of the @xmath57 components of the function @xmath44 with respect to the two variables @xmath58 and @xmath59 , and this derivative is taken at the point @xmath60 , substituting the expansion ( [ texpand ] ) into ( [ et ] ) gives e ( ) = | + o(^2 ) [ tbias ] since @xmath61 .
hence , the bias of the estimate @xmath44 is of order @xmath62 .
it easily follows from the expansion ( [ texpand ] ) that the covariance matrix of the estimate @xmath44 is given by @xmath63 \theta_i^t + { \cal o}(\sigma^4)\ ] ] ( it is not hard to see that the cubical terms @xmath64 vanish because the normal random variables with zero mean also have zero third moment , see also @xcite ) .
now , for the cartesian model @xmath65 = \sigma^2 i\ ] ] and for the radial model @xmath65 = \sigma^2 { \bf n}_i { \bf n}_i^t\ ] ] where @xmath25 is a unit normal vector to the curve @xmath14 at the point @xmath30
. then we obtain _
= ^2 _ i=1^n _ i _ i _ i^t + o(^4 ) [ csig0 ] where @xmath66 for the cartesian model and @xmath67 for the radial model . +
* lemma*. _ we have @xmath68 for each @xmath2 .
hence , for both models , cartesian and radial , the matrix @xmath69 is given by the same expression : _
_ = ^2 _ i=1^n _ i _ i^t + o(^4 ) [ csig ] this lemma is proved in appendix
. our next goal is now to find a lower bound for the matrix _
1:= _ i=1^n _ i_i^t [ calc1 ] following @xcite , we consider perturbations of the parameter vector @xmath70 and the true points @xmath71 satisfying two constraints .
first , since the true points must belong to the true curve , @xmath72 , we obtain , by the chain rule , _ * x * p(|*x*_i;| ) , @xmath73 stands for the scalar product of vectors .
second , since the identity ( [ tass ] ) holds for all @xmath3 , we get _ i=1^n _ i |*x*_i = [ tcon2 ] by using the notation ( [ ti ] ) .
now we need to find a lower bound for the matrix ( [ calc1 ] ) subject to the constraints ( [ tcon1 ] ) and ( [ tcon2 ] ) .
that bound follows from a general theorem in linear algebra : + * theorem ( linear algebra)*. _ let @xmath74 and @xmath75 .
suppose @xmath45 nonzero vectors @xmath76 and @xmath45 nonzero vectors @xmath77 are given , @xmath43 .
consider @xmath78 matrices @xmath79 for @xmath43 , and @xmath80 matrix @xmath81 assume that the vectors @xmath82 span @xmath83 ( hence @xmath84 is nonsingular ) .
we say that a set of @xmath45 matrices @xmath85 ( each of size @xmath78 ) is * proper * if _
a_i w_i = r [ propera1 ] for any vectors @xmath86 and @xmath87 such that u_i^tw_i + v_i^tr = 0 [ propera2 ] for all @xmath43 .
then for any proper set of matrices @xmath85 the @xmath80 matrix @xmath88 is bounded from below by @xmath89 in the sense that @xmath90 is a positive semidefinite matrix .
the equality @xmath91 holds if and only if @xmath92 for all @xmath2 .
_ + this theorem is , probably , known , but we provide a full proof in appendix , for the sake of completeness . as a direct consequence of the above theorem
we obtain the lower bound for our matrix @xmath93 : + * theorem ( kanatani - cramer - rao lower bound)*. _ we have @xmath94 , in the sense that @xmath95 is a positive semidefinite matrix , where _ _ ^-1 = _ i=1^n [ dmin ] in view of ( [ csig ] ) and ( [ calc1 ] ) , the above theorem says that the lower bound for the covariance matrix @xmath96 is , to the leading order , _ _
= ^2 d _ [ rc ] the standard deviations of the components of the estimate @xmath44 are of order @xmath97 . therefore
, the bias of @xmath44 , which is at most of order @xmath62 by ( [ tbias ] ) , is infinitesimally small , as @xmath47 , compared to the standard deviations .
this means that the estimates satisfying ( [ tass ] ) are practically unbiased .
the bound ( [ rc ] ) was first derived by kanatani @xcite for the cartesian functional model and strictly unbiased estimates of @xmath3 , i.e. satisfying @xmath98
. one can easily derive ( [ tass ] ) from @xmath99 by taking the limit @xmath100 , hence our results generalize those of kanatani .
here we derive an explicit formula for the covariance matrix of the weighted algebraic fit ( [ fmain3 ] ) and describe the weights @xmath36 for which the fit is statistically efficient . for brevity ,
we write @xmath101 .
we assume that the weight function @xmath102 is regular , in particular has bounded derivatives with respect to @xmath3 , the next section will demonstrate the importance of this condition .
the solution of the minimization problem ( [ fmain3 ] ) satisfies p_i^2 _ w_i + 2 w_i p_i _
p_i = 0 [ weq ] observe that @xmath103 , so that the first sum in ( [ weq ] ) is @xmath104 and the second sum is @xmath105 .
hence , to the leading order , the solution of ( [ weq ] ) can be found by discarding the first sum and solving the reduced equation w_i p_i _
p_i = 0 [ weq1 ] more precisely , if @xmath106 and @xmath107 are solutions of ( [ weq ] ) and ( [ weq1 ] ) , respectively , then @xmath108 , @xmath109 , and @xmath110 .
furthermore , the covariance matrices of @xmath106 and @xmath107 coincide , to the leading order , i.e. @xmath111 as @xmath47 . therefore , in what follows , we only deal with the solution of equation ( [ weq1 ] ) . to find the covariance matrix of @xmath44 satisfying ( [ weq1 ] ) we put @xmath112 and @xmath113 and obtain , working to the leading order , @xmath114 hence @xmath115^{-1 } \left [ \sum w_i ( \nabla_{\bf x } p_i)^t \ , ( \delta { \bf x}_i)\ , ( \nabla_{\theta } p_i)\right ] + { \cal o}(\sigma^2)\ ] ] the covariance matrix is then @xmath116\\ & = & \sigma^2 \left [ \sum w_i ( \nabla_{\theta } p_i ) ( \nabla_{\theta } p_i)^t \right ] ^{-1 } \left [ \sum w_i^2 \|\nabla_{\bf x } p_i\|^2 ( \nabla_{\theta } p_i ) ( \nabla_{\theta } p_i)^t \right ] \\ & & \times \left [ \sum w_i ( \nabla_{\theta } p_i ) ( \nabla_{\theta } p_i)^t \right ] ^{-1 } + { \cal o}(\sigma^3)\end{aligned}\ ] ] denote by @xmath117 the principal factor here , i.e.@xmath118^{-1 } \left [ \sum w_i^2 \|\nabla_{\bf x } p_i\|^2 ( \nabla_{\theta } p_i ) ( \nabla_{\theta } p_i)^t \right ] \
, \left [ \sum w_i ( \nabla_{\theta } p_i ) ( \nabla_{\theta } p_i)^t \right ] ^{-1}\ ] ] the following theorem establishes a lower bound for @xmath117 : + * theorem*. _ we have @xmath119 , in the sense that @xmath120 is a positive semidefinite matrix , where @xmath121 is given by ( [ dmin ] ) .
the equality @xmath122 holds if and only if @xmath123 for all @xmath2 . in other words ,
an algebraic fit ( [ fmain3 ] ) is * statistically efficient * if and only if the weight function @xmath124 satisfies w(x , y ; ) = [ wopt ] for all triples @xmath125 such that @xmath0 . here
@xmath126 may be an arbitrary function of @xmath3 . _
+ the bound @xmath127 here is a particular case of the previous theorem .
it also can be obtained directly from the linear algebra theorem if one sets @xmath128 , @xmath129 , and @xmath130^{-1 } ( \nabla_{\theta } p_i ) \ , ( \nabla_{\bf x } p_i)^t\ ] ] for @xmath43 . the expression ( [ wopt ] )
characterizing the efficiency , follows from the last claim in the linear algebra theorem .
here we illustrate our conclusions by the relatively simple problem of fitting circles . the canonical equation of a circle is ( x - a)^2 + ( y - b)^2 -r^2=0 [ circ0 ] and we need to estimate three parameters @xmath10 . the simple algebraic fit ( [ fmain2 ] ) takes form _ 2(a , b , r ) = _ i=1^n [ ( x_i - a)^2 + ( y_i - b)^2 -r^2]^2 [ f2 ] and the weighted algebraic fit ( [ fmain3 ] ) takes form _ 3(a , b , r ) = _ i=1^n w_i [ ( x_i - a)^2 + ( y_i - b)^2 -r^2]^2 [ f3 ] in particular , the graf becomes _
4(a , b , r ) = _
i=1^n [ f4 ] ( where the irrelevant constant factor of 4 in the denominator is dropped ) . in terms of ( [ dmin ] )
, we have @xmath131 and @xmath132 , hence @xmath133=4r^2\ ] ] therefore , _ = ( ccc u_i^2 & u_iv_i & u_i + u_iv_i & v_i^2 & v_i + u_i & v_i & n + ) ^-1 [ dmincir ] where we denote , for brevity , @xmath134 the above expression for @xmath121 was derived earlier in @xcite .
now , our theorem in section [ secse ] shows that the weighted algebraic fit ( [ f3 ] ) is statistically efficient if and only if the weight function satisfies @xmath135 .
since @xmath136 may be an arbitrary function , then the denominator @xmath137 here is irrelevant .
hence , statistically efficiency is achieved whenever @xmath138 is simply independent of @xmath5 and @xmath6 for all @xmath139 lying on the circle . in particular , the graf ( [ f4 ] ) is statistically efficient because @xmath140^{-1}=r^{-2}$ ] .
the simple af ( [ f2 ] ) is also statistically efficient since @xmath141 .
we note that the graf ( [ f4 ] ) is a highly nonlinear problem , and in its exact form ( [ f4 ] ) is not used in practice . instead
, there are two modifications of graf popular among experimenters .
one is due to chernov and ososkov @xcite and pratt @xcite : _
4(a , b , r ) = r^-2_i=1^n [ ( x_i - a)^2 + ( y_i - b)^2 -r^2]^2 [ f4a ] ( it is based on the approximation @xmath142 ) , and the other due to agin @xcite and taubin @xcite : _ 4(a , b , r ) = _ i=1^n [ ( x_i - a)^2 + ( y_i - b)^2 -r^2]^2 [ f4b ] ( here one simply averages the denominator of ( [ f4 ] ) over @xmath143 ) .
we refer the reader to @xcite for a detailed analysis of these and other circle fitting algorithms , including their numerical implementations .
we have tested experimentally the efficiency of four circle fitting algorithms : the olsf ( [ fmain1 ] ) , the simple af ( [ f2 ] ) , the pratt method ( [ f4a ] ) , and the taubin method ( [ f4b ] ) .
we have generated @xmath144 points equally spaced on a circle , added an isotropic gaussian noise with variance @xmath62 ( according to the cartesian model ) , and estimated the efficiency of the estimate of the center by e = [ e ] here @xmath145 is the true center , @xmath146 is its estimate , @xmath147 denotes averaging over many random samples , and @xmath148 , @xmath149 are the first two diagonal entries of the matrix ( [ dmincir ] ) .
table 1 shows the efficiency of the above mentioned four algorithms for various values of @xmath150 .
we see that they all perform very well , and indeed are efficient as @xmath49 .
one might notice that the olsf slightly outperforms the other methods , and the af is the second best .
[ cols=">,^,^,^,^",options="header " , ] table 3 .
data are sampled along a quarter of a circle .
it is interesting to test smaller circular arcs , too .
figure 1 shows a color - coded diagram of the efficiency of the olsf and the af for arcs from @xmath151 to @xmath152 and variable @xmath41 ( we set @xmath153 , where @xmath154 is the height of the circular arc , see fig . 2 , and @xmath155 varies from 0 to 0.5 ) .
the efficiency of the pratt and taubin is virtually identical to that of the olsf , so it is not shown here .
we see that the olsf and af are efficient as @xmath49 ( both squares in the diagram get white at the bottom ) , but the af loses its efficiency at moderate levels of noise ( @xmath156 ) , while the olsf remains accurate up to @xmath157 after which it rather sharply breaks down .
@xmath158 @xmath158 figure 1 : the efficiency of the simple olsf ( left ) and the af ( center ) .
the bar on the right explains color codes .
the following analysis sheds more light on the behavior of the circle fitting algorithms .
when the curvature of the arc decreases , the center coordinates @xmath159 and the radius @xmath160 grow to infinity and their estimates become highly unreliable . in that case the circle equation ( [ circ0 ] ) can be converted to a more convenient algebraic form a(x^2+y^2 ) + bx + cy + d = 0 [ abcd ] with an additional constrain on the parameters : @xmath161 .
this parametrization was used in @xcite , and analyzed in detail in @xcite .
we note that the original parameters can be recovered via @xmath162 , @xmath163 , and @xmath164 .
the new parametrization ( [ abcd ] ) is safe to use for arcs with arbitrary small curvature : the parameters @xmath165 remain bounded and never develop singularities , see @xcite .
even as the curvature vanishes , we simply get @xmath166 , and the equation ( [ abcd ] ) represents a line @xmath167 .
figure 2 : the height of an arc , @xmath154 , and our formula for @xmath41 . in terms of the new parameters @xmath165 ,
the weighted algebraic fit ( [ fmain3 ] ) takes form _ 3(a , b , c , d ) = _
i=1^n w_i [ a(x^2+y^2 ) + bx + cy + d]^2 [ ff3 ] ( under the constraint @xmath161 ) . converting the af ( [ f2 ] ) to the new parameters gives _
2(a , b , c , d ) = _
i=1^n a^-2 [ a(x^2+y^2 ) + bx + cy + d]^2 [ ff2 ] which corresponds to the weight function @xmath168 .
the pratt method ( [ f4a ] ) turns to _ 4(a , b , c , d ) = _ i=1^n [ a(x^2+y^2 ) + bx + cy + d]^2 [ ff4 ] we now see why the af is unstable and inaccurate for arcs with small curvature : its weight function @xmath168 develops a singularity ( it explodes ) in the limit @xmath169 .
recall that , in our derivation of the statistical efficiency theorem ( section 3 ) , we assumed that the weight function was regular ( had bounded derivatives ) .
this assumption is clearly violated by the af ( [ ff2 ] ) . on the contrary ,
the pratt fit ( [ ff4 ] ) uses a safe choice @xmath170 and thus behaves decently on arcs with small curvature , see next .
@xmath158 @xmath158 figure 3 : the efficiency of the simple af ( left ) and the pratt method ( center ) .
the bar on the right explains color codes .
figure 3 shows a color - coded diagram of the efficiency of the estimate of the parameter , hence the estimation of @xmath171 is equivalent to that of the curvature , an important geometric parameter of the arc .
] @xmath171 by the af ( [ ff2 ] ) versus pratt ( [ ff4 ] ) for arcs from @xmath151 to @xmath152 and the noise level @xmath153 , where @xmath154 is the height of the circular arc and @xmath155 varies from 0 to 0.5 . the efficiency of the olsf and the taubin method is visually indistinguishable from that of pratt ( the central square in fig .
3 ) , so we did not include it here .
we see that the af performs significantly worse than the pratt method for all arcs and most of the values of @xmath155 ( i.e. , @xmath41 ) .
the pratt s efficiency is close 100% , its lowest point is 89% for @xmath152 arcs and
@xmath172 ( the top right corner of the central square barely gets grey ) .
the af s efficiency is below 10% for all @xmath173 and almost zero for @xmath174 .
still , the af remains efficient as @xmath49 ( as the tiny white strip at the bottom of the left square proves ) , but its efficiency can be only counted on when @xmath41 is extremely small .
our analysis demonstrates that the choice of the weights @xmath36 in the weighted algebraic fit ( [ fmain3 ] ) should be made according to our theorem in section 3 , and , in addition , one should avoid singularities in the domain of parameters .
here we prove the theorem of linear algebra stated in section [ seckcr ] .
for the sake of clarity , we divide our proof into small lemmas : * lemma 2*. _ if a set of @xmath45 matrices @xmath85 is proper , then rank@xmath180 .
furthermore , each @xmath181 is given by @xmath182 for some vector @xmath183 , and the vectors @xmath184 satisfy @xmath185 where @xmath21 is the @xmath80 identity matrix .
the converse is also true . _
_ proof_. let vectors @xmath186 and @xmath187 satisfy the requirements ( [ propera1 ] ) and ( [ propera2 ] ) of the theorem . consider the orthogonal decomposition @xmath188 where @xmath189 is perpendicular to @xmath190 , i.e. @xmath191 . then the constraint ( [ propera2 ] )
can be rewritten as c_i = - [ propera3 ] for all @xmath2 and ( [ propera1 ] ) takes form _
i=1^n c_ia_iu_i + _
i=1^n a_iw_i^= r [ propera4 ] we conclude that @xmath192 for every vector @xmath189 orthogonal to @xmath190 , hence @xmath181 has a @xmath193-dimensional kernel , so indeed its rank is zero or one .
if we denote @xmath194 , we obtain @xmath195 .
combining this with ( [ propera3])-([propera4 ] ) gives @xmath196 since this identity holds for any vector @xmath87 , the expression within parentheses is @xmath197 .
the converse is obtained by straightforward calculations .
lemma is proved . *
lemma 3*. _ the sets of proper matrices make a linear variety , in the following sense .
let @xmath200 and @xmath201 be two proper sets of matrices , then the set @xmath85 defined by @xmath202 is proper for every @xmath203 .
_ _
proof_. for each @xmath2 consider the @xmath213 matrix @xmath214 .
using the previous lemma gives @xmath215 by construction , this matrix is positive semidefinite .
hence , the following matrix is also positive semidefinite : @xmath216 by sylvester s theorem , the matrix @xmath217 is positive semidefinite . _
proof_. assume that there is a proper set of matrices @xmath200 , different from @xmath221 , for which @xmath91 .
denote @xmath222 . by lemma 3 ,
the set of matrices @xmath223 is proper for every real @xmath224 . consider the variable matrix @xmath225 [ a_i(\gamma)]^t\\ & = & \sum_{i=1}^n a_i^{\rm o}(a_i^{\rm o})^t + \gamma\left ( \sum_{i=1}^n a_i^{\rm o}(\delta a_i)^t + \sum_{i=1}^n ( \delta a_i)(a_i^{\rm o})^t\right ) + \gamma^2\sum_{i=1}^n ( \delta a_i)(\delta a_i)^t\end{aligned}\ ] ] note that the matrix @xmath226 is symmetric . by lemma 5
we have @xmath227 for all @xmath224 , and by lemma 6 we have @xmath228 .
it is then easy to derive that @xmath229 .
next , the matrix @xmath230 is symmetric positive semidefinite . since we assumed that @xmath231 , it is easy to derive that @xmath232 as well .
therefore , @xmath233 for every @xmath2 .
the theorem is proved .
g. taubin , estimation of planar curves , surfaces and nonplanar space curves defined by implicit equations , with applications to edge and range image segmentation , _ ieee transactions on pattern analysis and machine intelligence _ , * 13 * , 1991 , 11151138 . | we study the problem of fitting parametrized curves to noisy data . under certain assumptions ( known as cartesian and radial functional models )
, we derive asymptotic expressions for the bias and the covariance matrix of the parameter estimates .
we also extend kanatani s version of the cramer - rao lower bound , which he proved for unbiased estimates only , to more general estimates that include many popular algorithms ( most notably , the orthogonal least squares and algebraic fits ) .
we then show that the gradient - weighted algebraic fit is statistically efficient and describe all other statistically efficient algebraic fits .
keywords : least squares fit , curve fitting , circle fitting , algebraic fit , rao - cramer bound , efficiency , functional model . |
since graphene was discovered , many interesting physical effects were predicted and some were experimentally confirmed.@xcite graphene nowadays is regarded as the new - generation material and draws extensive attentions .
the pn junction is the simplest structure in graphene system and is the brick to build up more complicated structures such as pn superlattice,@xcite quantum dot,@xcite and point contact.@xcite lots of intriguing effects are induced by the pn junction , for examples , klein tunneling for normally injected electrons,@xcite negative refraction and veselago lens for electron beams,@xcite , snake - hall states around the interface under strong magnetic field,@xcite , and non - integer conductance quantization due to the mix of electron and hole landau edge modes near the pn interface.@xcite in ref .
[ ] , an analytical formula for the graphene pn conductance was proposed for the special case that the fermi energy lies at the middle point between the potentials at two ends .
however , the analytical calculation of pn junctions for the whole bipolar region has been lacked for a long time .
furthermore , spin - orbit coupling ( soc ) was not included in the calculation of ref .
[ ] , which plays more and more important role in nowadays physics society .
it is well known that the soc is too weak to be observed in pristine graphene.@xcite however , the soc deserves to be considered because it can be dramatically magnified to measurable scale by a few methods.@xcite moreover , a family of graphene - like materials such as silicene , germanene , and stanene ( single layer of silicon , germanium , and stannum ) were recently reported to have the same honeycomb lattice as graphene but two types of sub - lattices buckled up and down out of plane .
the lattice buckling leads to observable soc,@xcite which make these graphene and graphene - like materials be topological insulators.@xcite by including the soc and stagger potential , the graphene lattice model is not a problem for graphene itself , but a platform to study different types of junctions@xcite and varies phases of two - dimensional topologic insulators.@xcite ( a ) sketch of the tunneling when electrons pass the pn junction .
( b ) the area of enclosed by @xmath0 in the @xmath1-@xmath0 plane in the process of an electron transmiting through the pn junction.,width=264 ] motivated by these reasons , we studied the conductance of gradual graphene pn junctions with the soc considered , which induces a energy gap and drives graphene to be a topologic insulator . by means of the subband tunneling model@xcite , we derived an analytical conductance expression of the graphene pn junction in the whole interval of the bipolar region . to check the validity of our analytical result
, we numerically calculated the pn conductance using the non - equilibrium green s function ( negf ) method , and found the conductances obtained by the two methods fit each other strikingly well . according to the analytical result
, there exists a parameter interval , in which the conductance drops a constant value with respect to that without gap . when the stagger potential is present , the energy gaps for valleys @xmath2 and @xmath3 become different . the analytical expression allows us to decomposes the conductance into valley @xmath2 and @xmath3 components .
the conductance is always governed by the valley having the smaller gap .
the conductance versus the stagger potential consists of two gaussian peaks , which are contributed by the two valleys .
this effect can be used for valley detection in the graphene - like materials without using double - interface resonant devices .
the dispersion of graphene or graphene - like lattice around one valley with a uniform background potential @xmath4 reads @xmath5 where @xmath6 is the reduced plank constant , @xmath7 is the fermi velocity , and @xmath8 , @xmath9 , @xmath10 and @xmath4 are the wave - vector , energy gap , energy of electron , and background potential .
here we has not specified the physical origin of the energy gap , it can be induced by stagger potential , soc , or the combination of both .
for the electrons in a ribbon along @xmath1-direction , @xmath1-component wave - vector is a good quantum number , and the motion in @xmath11-direction is described by lateral mode index @xmath12 .
the dispersion relation of a graphene ribbon is @xmath13 where @xmath14 is the lateral energy of mode @xmath12 , which is positive for electrons and negative for holes , and @xmath15 is longitudinal wave - vector of electron of subband @xmath12 . in the equation , @xmath16 does not appear because its effect is absorbed in @xmath14 .
the minimum or maximum energies of the electron or hole dispersions for a given mode @xmath12 is regarded as the subband edge , which is obtained by letting @xmath17 as @xmath18 when the ribbon is subjected by a pn junction potential @xmath19 , the uniform potential @xmath4 in eqs .
( [ ribbon ] ) and ( [ edge ] ) is replaced by @xmath19 , and the subband edges are modulated by the pn potential , as illustrated in fig [ area ] ( a ) . in this case @xmath15 is no longer a good quantum number and varies when electron propagating , saying , @xmath20 .
now we track the physics when the electron goes through the pn junction with the help of fig .
[ area ] ( a ) .
we consider an electron - like particle of lateral mode @xmath12 ( @xmath21 in the figure ) starts from @xmath22 .
before it is injected in the pn region , it has a real wave - vector @xmath15 . after it enters the pn area , the subband edge @xmath23 increases , @xmath15 decrease and goes into zero at the turning point @xmath24 .
after it moves beyond the point , @xmath15 becomes imaginary ( the electron goes evanescently with the evanescent wave - vector @xmath25 ) , and @xmath26 becomes more and more negative .
this indicates @xmath27 in eq .
( [ ribbon ] ) decreases continuously , and runs into zero at point @xmath28
. the electron can no longer travel on of mode @xmath12 and has to transit to subband @xmath29 as a hole , i.e. , from point @xmath28 to @xmath30 .
the evanescent wave propagation continues until the electron reaches the turning point @xmath31 on the curve of subband edge @xmath29 , the wave - vector recovers to a real value , and propagates away .
according to the wkb theory , the transmission is determined by the evanescent transport part as @xmath32 where @xmath33 stands for the area enclosed by @xmath34 and @xmath35 in the @xmath1-@xmath0 plane .
we can approximate the curves of @xmath36 and @xmath37 on the @xmath1-@xmath0 plane as two quarters of two different ellipses @xcite sharing one principle radius @xmath38 .
the other radius of each quarter is determined by the explicit form of @xmath19 .
the two ellipse quarters are shown in fig .
[ area ] ( b ) .
letting @xmath39 be the distance between @xmath24 and @xmath31 in @xmath1-direction , the area of the two quarters reads @xmath40 the conductance of the pn junction is thus the summation of the contributions of all the subband involved into the pn tunneling .
( a ) conductance @xmath41 ( in units of @xmath42 ) for @xmath43 versus @xmath44 , where @xmath45 is the in - plane atom - atom distance .
( b ) @xmath46 ( @xmath41 at @xmath47 ) in eq .
( [ g0 ] ) versus @xmath48 .
( c)-(e ) numerical calculated @xmath46 ( in units of @xmath42 ) versus @xmath49 ( in units of @xmath50 ) .
parameters are @xmath51 and @xmath52 with @xmath53 the nearest hopping energy in tight - binding model . ,
width=302 ] the theory of the transmission for a pn junction of a ribbon can be applied to the bulk system .
given an electron injected with the wave - vector @xmath54 , the lateral energy @xmath14 should be replaced with @xmath55 , where plus for electrons and minus for holes .
the distance between the turning points @xmath24 and @xmath31 ( see fig . [ area ] ) depends on not only the lateral energy but the potential slope at the incident energy @xmath56 after substituting @xmath39 by @xmath57 , we have the transmission @xmath58,\end{aligned}\ ] ] where @xmath48 is a parameter of dimension of energy @xmath59 the left and right end ( n - end and p - end ) are assumed to be lifted by @xmath60 and @xmath61 respectively .
if we set @xmath62 and @xmath63 , by defining the effective length @xmath64 and the incident angle @xmath65 , we recover the formula appearing in ref .
[ ] , @xmath66 . by counting all the transmission of different @xmath67
, we have the conductance across the pn juntion , @xmath68 where @xmath69 is the width of the sample and erf represents the error function . in the equation
, it has to bear in mind that @xmath48 is a function of energy , saying , @xmath70 . to examine the validation of eq .
( [ g ] ) , we numerically calculated conductance of graphene pn junctions for comparison .
the pn potential is modeled by a half period of sinusoidal function of the length @xmath49 connecting the two ends .
the numerical conductance is obtained by the negf method based on the tight - bonding hamiltonian including the kane - mele soc .
the soc induces an energy gap @xmath71 , where @xmath72 is a parameter to characterize the soc amplitude in the tight - binding hamiltonian . to eliminate the boundary effect and simulate the infinite system ,
the periodical edge condition is adopted .
the details of numerical negf calculation and the tight - binding hamiltonian are not shown here but presented in the appendix for completeness .
figure [ betal ] ( a ) shows the conductances calculated from numerical and analytical methods . as one can see , they are consistent very well .
these conductance curves are peaked and centered at @xmath47 with peak width @xmath73 .
larger energy gap means larger area of ellipses shown in fig .
[ area ] ( b ) , and leads to smaller conductance .
( [ g ] ) is obtained from the wkb approximation , so the consistency will break down for the potential slope larger than a criterion value @xmath74 . by comparison the conductance for many choices of @xmath49
, we estimate the criterion slope to be @xmath75 , where @xmath53 is the nearest hopping energy and @xmath45 is the in - plane atom - atom distance . for the graphene lattice , @xmath76ev , @xmath77
, so we have @xmath78v/@xmath79 m , and similar values of @xmath74 for other graphene - like materials , much greater than the experimentally achieved value , which is only several v/@xmath79m.@xcite in realistic situations , the slope of pn potentials always allow the wkb approximation to be applied safely .
particularly , if the fermi energy is aligned at the middle energy between the potentials of two ends , saying , @xmath47 , the conductance reaches its maximum @xmath80 in the equation and from here on , we omit the constant pre - factor @xmath81 for simplicity . because the potential varies slowly along @xmath1 , we consider the situation @xmath82 . for this case
the error function gives 1 , and the conductance is reduced to @xmath83 when no bulk gap exists , saying , @xmath62 , it is interesting to note the conductance is simply proportional to @xmath48 .
figure [ betal ] ( b ) shows the conductance at @xmath47 as a function of the junction length .
the conductance remains almost non - changed when @xmath48 below a threshold , and then changes linearly with @xmath48 increases . in this special linear region ,
the conductance reads @xmath84 when @xmath48 is out of the region , say , when much smaller than the lower limit , no conductance can be observed , and when much larger than the upper limit , the exponential term in eq .
( [ g0 ] ) tends to unity and the conductance is proportional to @xmath48 .
equation ( [ g0para ] ) means that the conductance of a system with an energy gap is smaller than that of gapless system by a constant value in the parameter region specified in the equation .
this interval of @xmath48 can be translated to a scope of the junction length as , @xmath85 where @xmath86 .
figures [ betal ] ( c ) through ( e ) show numerically calculated @xmath46 versus @xmath49 for both the cases @xmath62 and @xmath87 .
for the case of @xmath52 and @xmath88 , we have @xmath89 . in the interval of @xmath49 from @xmath90 through @xmath91 ,
the difference between the two conductances is almost a constant , as illustrated in fig .
[ betal ] ( d ) .
conductance and its valley components ( in units of @xmath42 ) as functions of @xmath44 ( in units of @xmath53 ) for ( a ) @xmath43 and ( b ) @xmath92 .
conductance and its valley components ( in units of @xmath42 ) at @xmath47 as functions of @xmath93 ( in units of @xmath53 ) for ( c ) @xmath43 and ( d ) @xmath92 .
the curves of total conductance are calculated numerically and the conductance components of valleys are calculated analytically .
other parameters are @xmath94 and @xmath52 .
, width=302 ] if there is a stagger potential present , saying , potentials on a - type and b - type atoms are lifted by @xmath93 and @xmath95 respectively .
the stagger potential leads to the electron dispersion and the hole dispersion come closer for one valley and leave apart for the other valley .
thus , the energy gaps of the two valleys are driven to be different , say , @xmath96 the valley dependent energy gap results in that the conductances at @xmath47 for different valley are @xmath97.\end{aligned}\ ] ] the total conductance is @xmath98 .
this equation reveals that when we treat the stagger potential as a tunable parameter , the conductance component of a single valley as a function of the stagger potential is a gaussian peak , which is centered at @xmath99 or @xmath100 , depending on the valley , and the full width at half maximum ( half - width ) of which is @xmath101 .
the peak phenomenon of the conductance has two important features .
first , the maximum conductance is proportional to @xmath48 , and for a given pn potential height @xmath61 , it is proportional to @xmath102 .
so the peak will survive for super long pn junctions and one need not to worry about the conductance signal fades out .
second , the peak feature is not a resonant effect , and need not tune the parameters to meet the resonant condition .
figure [ ged ] ( a ) through ( d ) shows the conductance and the conductance components of valleys @xmath2 and @xmath3 of pn junctions as functions of the fermi energy and as functions of of stagger potential amplitude @xmath93 .
the conductance contributed by both valleys is calculated numerically , and the conductance component of each valley is analytically . in fig .
[ ged ] ( a ) , it can be clearly seen that the conductance consists of two peaks with different peak widths .
the two peaks are contributed by valleys @xmath2 and @xmath3 , and the valley with the smaller gap leads to the wider and higher peak . for a longer pn junction ,
the decrease of the conductance component of valley @xmath3 is more notably than that of valley @xmath2 , as shown in fig .
[ ged ] ( b ) .
the conductance is always dominated by the transport of the valley with the smaller gap . in fig .
[ ged ] ( c ) , the conductance curve shows a peak with maximum at @xmath103 , but it indeed consists of two peaks of gaussian type . for the parameters in fig .
[ ged ] ( c ) , the half - width is about @xmath104 , greater than the peak separation @xmath105 , and the two peaks can not be distinguished on the conductance curve .
figure [ ged ] ( d ) shows the conductance and components for a longer pn junction . since longer @xmath49 means smaller @xmath48 and thus narrower peaks ( half - width is about @xmath106 , smaller than the peak separation @xmath107 ) , the resolved peaks appear on the total conductance curve .
one can choose a longer @xmath49 so as to the two peaks are totally resolved .
all the above discussions are based on single spin .
indeed the conductance for the other spin is the same , i.e. , @xmath108 . because flipping the spin leads to the interchange of the gaps of valleys @xmath2 and @xmath3 , the conductance component of valley @xmath2 for spin - up and that of valley @xmath3 for spin - down are equivalent , i.e. , @xmath109 .
the peak - behavior of the conductance can be used to detect valley index of electric current and to measure the valley - polarization by combining a pn junction with a spin - filter . freely tuning
the stagger potential may be a challenge for graphene , but is easy for other graphene - like materials such as silicene , germanene and stanene , due to the lattice buckling .
many schemes of valley detection proposals in other literatures rely on valley - dependent resonance of double - interface scattering , which needs accurate parameter control and is difficult to result in reliable observation .
we obtained an analytical expression of the conductance of graphene pn junctions in the bipolar region with the soc included and verified its validity by comparing it and the numerically calculated conductance .
our analytical result indicates that the conductance as a function of stagger potential consists of two gaussian peaks , which are contributed by the tunneling of electrons in two valleys respectively .
this effect can be used for valley detection .
this work was supported by nsf of china grant no .
11274124 , no . 11474106 , no .
11174088 , and hebei nsf grant no .
the tight - binding hamiltonian of a perfect graphene ribbon including the soc reads @xmath110 where @xmath111 ( @xmath112 ) is the creation ( annihilation ) operator for an electron with spin @xmath113 on site @xmath114 , @xmath115 is the @xmath116-component of pauli matrix , and the summations with the brackets @xmath117 and @xmath118 run over all the nearest and next - nearest neighbor sites , respectively .
the first term is the hamiltonian of the nearest neighbor hopping with hopping energy @xmath53 .
the second term is the soc hamiltonian which involves the next - nearest neighbor hopping with amplitude @xmath119 and a path dependent factor @xmath120 . for the electron couples form atom @xmath114 , mediated by a nearest neighbor site and to a next - nearest neighbor atom @xmath121
, we have @xmath122 if it makes a left turn and @xmath123 if goes a right turn .
the soc term results in an energy gap as @xmath124 the hamiltonian including the stagger potential , if it is present , reads @xmath125 where @xmath93 is the stagger potential amplitude , and @xmath126 when @xmath114 represents an a - type atom and @xmath127 for a b - type atom .
the stagger potential itself means a gap @xmath128 . the combination of the stagger gap and the soc gap results in two different gaps for valleys @xmath2 and @xmath3 , as in eq .
( [ deltak ] ) . when the ribbon is subjected by a space - varying potential ,
the hamiltonian is @xmath129 where @xmath130 is the space - varying potential on site @xmath114 .
the conductance can be calculated by the negf method as @xmath131 where @xmath132 and @xmath133 are the retarded ( advanced ) green s function of the area covered by the pn potential and the linewidth function of the left ( right ) lead .
all these functions are obtained through the tight - binding hamiltonian . | graphene pn junction is the brick to build up variety of graphene nano - structures .
the analytical formula of the conductance of graphene gradual pn junctions in the whole bipolar region has been absent up to now . in this paper , we analytically calculated that pn conductance with the spin - orbit coupling and stagger potential taken into account .
our analytical expression indicates that the energy gap causes the conductance to drop a constant value with respect to that without gap in a certain parameter region , and manifests that the curve of the conductance versus the stagger potential consists of two gaussian peaks
one valley contributes one peak .
the latter feature allows one to detect the valley polarization without using double - interface resonant devices . |
thank you to safa motesharrei for his corrections and comments .
was partially supported by nsf grant dms-0616585 and nih grant r01-hg0294501 .
e.s . was partially supported by nsf grants dms-0639300 and dms-0907818 , and nih grant r01-mh79502 . | the presence of a period - doubling cascade in dynamical systems that depend on a parameter is one of the basic routes to chaos .
it is rarely mentioned that there are virtually always infinitely many cascades whenever there is one .
we report that for one- and two - dimensional phase space , in the transition from no chaos to chaos as a parameter is varied there must be infinitely many cascades under some mild hypotheses .
our meaning of chaos includes the case of chaotic sets which are not attractors .
numerical studies indicate that this result applies to the forced - damped pendulum and the forced duffing equations , viewing the solutions once each period of the forcing .
we further show that in many cases cascades appear in pairs connected ( in joint parameter - state space ) by an unstable periodic orbit .
paired cascades can be destroyed or created by perturbations , whereas unpaired cascades are conserved under even significant perturbations . in fig .
[ fig : quadratic ] as @xmath0 increases towards a value @xmath1 , one encounters a family of periodic orbits that undergo an infinite sequence of period doublings with the period of these orbits tending to @xmath2 .
the appearance of infinitely - many such ( period - doubling ) cascades is one of the most prominent features observed in the study of parametrized maps .
cascades were first reported by myrberg in 1962 @xcite ( _ cf . _ fig .
[ fig : quadratic ] ) , is shown in blue .
there are infinitely - many cascades , each with infinitely - many period - doubling bifurcations .
each saddle - node bifurcation creates both a cascade and a path of unstable orbits , the latter shown as black curves for periods up to six .
we call the paths of unstable orbits the * stems * of the cascades .
below the largest gaps or _ windows _ in the blue regions , we give the period of the associated cascade .
for this particular map , the black curves of unstable orbits extend to @xmath3 . here , at each sufficiently large @xmath0 , there is exactly one orbit ( @xmath4 points ) in each period-@xmath4 stem .
[ fig : quadratic],title="fig : " ] [ fig : quadratic ] and robert may popularized their existence to a huge scientific audience @xcite .
they are found in a large variety of contexts : in raleigh - bernard convection @xcite , damped bouncing balls @xcite , van der pol oscillators @xcite ( _ cf .
_ fig .
[ fig : vdp ] ) , map of the double - well duffing equation : @xmath5 .
numerical studies show regions of chaos interspersed with regions without chaos , indicating that our off - on - off chaos theorem applies to this situation .
[ fig : vdp ] ] vibratory ball milling @xcite , dust charge fluctuation in plasma systems @xcite , external optical injection in lasers @xcite , delay oscillators @xcite , vibrating damaged structures @xcite , glow discharge @xcite , combustion @xcite , very slow classical cepheid stars @xcite , neuron and pancreatic cells @xcite , dripping faucets @xcite , bouncing droplets on soap films @xcite , belousov - zhabotinsky reactions @xcite , and in the cromorn , a medieval musical instrument @xcite .
we find that for many systems depending on a parameter , every periodic orbit is part of a cascade ; for such systems , cascades are as fundamental as periodic orbits themselves .
the scaling properties of individual cascades have been studied for cascades in nearly quadratic maps @xcite , but there are only a few results about the existence of cascades @xcite .
furthermore , the mathematical and scientific literature focuses on the study of single cascades . in this letter
, we describe the results of a new general theory of cascades , which explains why cascades exist and why chaotic dynamical systems often have infinitely many cascades . *
two kinds of periodic orbits meet at period - doubling bifurcations . * for a map @xmath6 that depends on a parameter @xmath0 , a point @xmath7 is a * period-@xmath8 point * if @xmath9 , where @xmath10 is chosen as small as possible . writing @xmath11 , its * ( periodic ) orbit *
is the set of points @xmath12 . by the * eigenvalues of * that orbit , we mean the eigenvalues of the jacobian matrix @xmath13 . if the map is one - dimensional , then by the derivative of the orbit , we mean the derivative of @xmath14 .
a * bifurcation orbit * is any orbit having an eigenvalue with absolute value @xmath15 .
we call a periodic orbit a * flip * orbit if the orbit has an odd number of eigenvalues less than -1 , and -1 is not an eigenvalue .
( in one dimension , this condition is : derivative @xmath16 . in dimension two ,
flip orbits are those with exactly one eigenvalue @xmath17 . )
all other periodic orbits are called * regular*. a family switches between flip and regular orbits at a period - doubling bifurcation orbit since this bifurcation orbit always has an eigenvalue equal to @xmath18 . for @xmath19 in some interval @xmath20 ,
let @xmath21 be a path of _ regular _ periodic points in @xmath22 space and write @xmath23 $ ] to denote the periodic orbit of the periodic point @xmath24 .
we say @xmath25)$ ] is a * cascade * if the path contains infinitely many period - doubling bifurcations and has orbits with all the periods @xmath26 for some positive integer @xmath4 . , where @xmath27 .
cascades @xmath28 and @xmath29 constitute a * pair * of cascades , since they are connected by a family of orbits ( black curves ) .
cascade @xmath30 is unbounded and unpaired , as indicated by the black curve continuing to the right to @xmath31 .
[ fig : bounded ] ] * assumptions .
* _ all results in this paper assume that the maps @xmath32 are infinitely differentiable jointly in the parameter @xmath0 and in phase space @xmath33 and that all of the periodic orbit bifurcations are * generic * , meaning that every bifurcation orbit is a standard nondegenerate period - doubling , saddle - node , or hopf bifurcation .
_ for example , generic maps have neither symmetries nor symmetry breaking bifurcations . nor
do they have pitchfork or period - tripling bifurcations .
we also assume that for any parameter interval @xmath34 $ ] , all periodic orbits are contained in a bounded set in phase space . * the cascades route to chaos in dimensions one and two . *
usually the term `` route to chaos '' refers to formation of _ chaotic attractors _ , in which case there are several possible routes to chaos . in this letter , we consider routes from a parameter having no chaos to one with _ chaotic sets _ , where the sets are not necessarily attractors . in fact , we require only one aspect of chaos : we say that there is * chaos * at a particular @xmath0 if there are infinitely many regular periodic orbits .
for example , there is chaos whenever there is a transverse homoclinic point .
that is equivalent to having a horseshoe for some iterate of the map , and it implies there are infinitely many regular saddles in two dimensions , and infinitely many regular unstable orbits in dimension one .
this definition of chaos is sufficiently general as to include having one or multiple coexisting chaotic attractors , as well as the case of transient chaos .
we say a map has * no chaos * at a particular @xmath0 if there are at most finitely many regular periodic orbits .
* cascades route to chaos theorem . *
_ assume the parametrized map @xmath35 has a one- or two - dimensional phase space .
if at parameter value @xmath36 there is no chaos , and at @xmath37 , there is chaos and there are at most finitely many attracting orbits ( and finitely many repelling orbits in dimension 2 ) , then @xmath38 has infinitely many cascades between @xmath36 and @xmath37 . _
thus for one- and two- dimensional smooth families , the only route to chaos is through infinitely many cascades .
it is for instance impossible to get chaos at @xmath37 from a single cascade . if we were to omit the assumption that there are infinitely many _
regular _ orbits , the conclusion would be false .
for example , in the quadratic map let @xmath39 denote the feigenbaum parameter value , i.e. , the first parameter value where the period doublings accumulate . at @xmath39
there are infinitely many periodic orbits , and all but one are flip orbits . furthermore @xmath39 is preceded by a single cascade .
each cascade will have some minimum period @xmath8 and will have periodic orbits of periods @xmath40 . if the map is one - dimensional or is two - dimensional and dissipative in the sense that there are no repelling periodic orbits , then each of these periods is the period of some attracting orbits in the path , though not all orbits will necessarily be attracting .
based on numerical studies , a number of maps appear to satisfy the conditions of the cascades route to chaos theorem .
note that these numerical verifications involve significantly more work than just plotting the attracting sets for each parameter , since we are concerned about both the stable and the unstable behavior to determine whether there is chaos .
examples include the time-@xmath41 maps for the double - well duffing ( fig .
[ fig : vdp ] ) , the triple - well duffing , and forced - damped pendulum ( fig .
[ fig : pend ] ) , as well as the ikeda map used to describe the field of a laser cavity , and the pulsed damped rotor map . were plotted in red for the time-@xmath41 map of the forced - damped pendulum equation : @xmath42 , indicating the general areas with chaotic dynamics for this map .
then the attracting sets were plotted in black , hiding some periodic points .
much more detailed calculations confirm that interspersed with the chaos , there are some parameter ranges without chaos .
[ fig : pend ] ] we now discuss the creation of a path @xmath43)$ ] of _ regular orbits_. consider the procedure of starting at a regular periodic orbit and following the path of periodic orbits containing it in @xmath44-space , following only regular orbits .
follow the path such as numerically through saddle - node bifurcations by reversing the direction in the parameter space . when period-@xmath8 orbits reach a period - doubling bifurcation @xmath45)$ ] ,
follow the period-@xmath46 orbits .
they are regular orbits , whereas the extension past @xmath45)$ ] of period-@xmath8 orbits are flip orbits .
likewise , at a period - halving bifurcation , follow along the branch of regular orbits .
following a path away from a cascade can lead to @xmath47 or it can lead to a second cascade .
we call an entire connected path of regular orbits a * component*. two orbits are in the same component if and only if it is possible to follow a path of regular orbits from one orbit to the other continuously , without jumps , though the period is allowed to double many times . * there are two types of cascades .
* only one type of cascades occurs for quadratic maps , but there is a second type of cascade as well .
namely , the component of a cascade either lies in a bounded region of @xmath44-space , or it does not .
accordingly we call the cascade * bounded * or * unbounded * , respectively , though the adjective really describes the cascade s component . each cascade in the quadratic map ( _ cf . _ fig .
[ fig : quadratic ] ) is unbounded because a branch of periodic orbits extends from each cascade to @xmath48 .
the constant period branch that extends to infinity is called the * stem * of the ( unbounded ) cascade . thus the theory of quadratic maps
is the theory of unbounded cascades .
we call the period of the orbits in the stem , the * stem period of the unbounded cascade*. ( we only treat the case of an unbounded cascade which has a constant period stem . )
we refer to two cascades as a * pair * if they are in the same component ; that is , if the two cascades have a path of periodic orbits running from one the other ( _ cf . _
fig .
[ fig : bounded ] ) .
we have shown the following rigorous result .
* bounded paired cascades theorem . * _ bounded cascades always come in pairs .
_ the theorem below shows that paired cascades are common .
once again we consider @xmath49 in a one- or two - dimensional phase space .
we say that map @xmath38 has * off - on - off chaos * for @xmath50 if there is no chaos at @xmath36 and @xmath51 , whereas at @xmath37 , @xmath38 is as assumed in the cascades route to chaos theorem
. * off - on - off chaos theorem . * _ if @xmath38 is two - dimensional and has off - on - off chaos for @xmath50 , then @xmath38 has infinitely many bounded paired cascades and at most finitely many unbounded cascades .
_ our numerical studies indicate that the time-@xmath41 maps of the forced double - well duffing ( fig .
[ fig : vdp ] ) and forced damped pendulum ( fig .
[ fig : pend ] ) have off - on - off chaos , and that it occurs on multiple non - overlapping parameter regions . here
@xmath41 is the forcing period .
in fact we can not prove there are only finitely many periodic attractors , though we find very few .
these systems have no periodic repellers .
* conservation of unbounded cascades . *
if in addition to the hypotheses of the cascades route to chaos theorem , we assume further that ( _ i _ ) at @xmath36 there are no regular periodic orbits and ( _ ii _ ) at @xmath37 there are no attractors ( nor repellers in dimension two ) , then the component of _ each _ regular orbit at @xmath37 contains a cascade that is between @xmath36 and @xmath37 , and distinct regular orbits at @xmath37 are in different components . starting at each regular orbit at @xmath37 , there is a unique path of regular orbits initially in the direction of @xmath36 ending in a cascade .
thus the number of cascades hitting the boundary of the parameter interval depends only on the number of regular orbits at @xmath37 , independent of the behavior of the map between @xmath36 and @xmath37 .
this leads to a heuristic * conservation principle for unbounded cascades * : if @xmath52 satisfies property ( _ i _ ) for @xmath0 sufficiently negative and property ( _ ii _ ) for @xmath0 sufficiently positive , then any perturbation @xmath53 such that @xmath38 dominates @xmath54 for @xmath47 has the same set of unbounded cascades .
three rigorous examples of this idea are encapsulated in the following one - dimensional maps @xmath55 where for some real positive @xmath56 , latexmath:[\[\begin{aligned } \label{eqn : g1 } for such @xmath54 , each of the three maps each has no regular periodic orbits for @xmath0 sufficiently negative , and for @xmath0 sufficiently large has a one - dimensional horseshoe map .
the quartic map was chosen so that when @xmath54 is identically 0 and @xmath58 , the graph has two minima , at @xmath59 , and the local maximum is at @xmath60 .
the conditions on @xmath54 guarantee that it does not significantly affect the periodic orbits when @xmath61 is sufficiently large ; in particular it does not affect their eigenvalues , so it does not affect the number of period-@xmath8 regular periodic orbits for large @xmath61 . hence one can check that all three maps have no regular periodic orbits for @xmath0 very negative and have no attracting periodic orbits for @xmath0 very positive , and for sufficiently large @xmath61 , all periodic orbits are contained in the set @xmath62 $ ] .
thus the following holds .
* conservation of cascades theorem . * _ for any @xmath38 of the functions in eqn .
[ eqn : quad+g ] , with @xmath54 chosen as in eqn .
[ eqn : g1 ] , the number of stem - period-@xmath4 unbounded cascades is independent of the choice of g. _ [ table1 ] . the number of unbounded cascades of stem period @xmath4 for any large - scale perturbation of the quadratic , cubic , and quartic maps , respectively labeled @xmath63 , @xmath64 , and @xmath65 . [
cols="<,<,<,<",options="header " , ] the number of unbounded cascades in each of these three cases is summarized in our table .
the relative sizes reflect the complexity of the three horseshoe maps ; the quadratic map has the complexity of a standard two - branch horseshoe , whereas the cubic has a horseshoe with three branches , and the quartic has four branches .
the conservation principle works for two- and higher - dimensional maps as well .
we have shown in @xcite that large - scale perturbations of the two - dimensional hnon map conserve unbounded cascades . in @xcite
, we have shown that there is conservation of unbounded cascades for a coupled system of @xmath66 quadratic maps . |
the heaviest post - main sequence mass - loss occurs during the red supergiant ( rsg , high mass stars ) phase or at the tip of the asymptotic giant branch ( agb , intermediate mass stars ) . in oxygen - rich environments , the velocity structure of the inner part of the circumstellar envelope below the dust formation
can be studied in sio maser emission ( chapman & cohen 1986 ) .
this is where the poorly understood initiation of the mass loss flow occurs .
recently , the first obscured evolved stars in the lmc were found , using the iras database ( reid 1991 ; wood et al .
1992 ; zijlstra et al .
wood et al .
( 1992 ) discovered oh maser emission from six of these stars : the first known extra - galactic oh / ir stars .
we used the sest at la silla to observe the best candidates for detection of 86 ghz sio maser emission among these oh / ir stars .
as a first result , we present the discovery of the first extra - galactic sio maser viz . the rsg iras045536825 .
the observations were performed on days 25 , 30 , and 31 of may 1995 .
we used the 15 m sest ( sub)mm telescope at eso / la silla with the new 3 mm sis receiver as the frontend , and the high resolution spectrograph hrs as the backend .
we observed the vibrationally excited rotational transition of the silicate molecule sio(21)@xmath0 at 86243.442 mhz . during part of the time we used this configuration simultaneously in combination with the low resolution spectrograph lrs tuned at the same frequency .
we observed in a dual beam switch mode with the source alternately placed in the two beams , a method which yields very flat baselines .
the beam separation was @xmath1 .
the pointing was checked every few hours , using the nearby bright sio maser r dor .
this agreed with the specified accuracy of @xmath2 rms .
the full width half power of the beam is @xmath3 . at 86
ghz the channel separation is 0.15 km s@xmath4 for the hrs , and 2.4 km s@xmath4 for the lrs .
the conversion factor from antenna temperature to flux units is 25 jy k@xmath4 at 86 ghz .
the internal absolute flux calibration is accurate to about 20% .
helped by good atmospheric conditions relative humidity 15 30% , and outside air temperature about 15@xmath5c with little or no cirrus the system temperature at 86 ghz was about 120 k or less at elevations above 40@xmath5 , and about 150 k at an elevation of 20@xmath5 .
polarization biases were minimized by observing at different hour angles , because the parallactic angle changes with hour angle .
long integrations naturally take care of this .
with the spectrograph having 1600 2000 channels though not all independent of each other a minimum value for a detection threshold is three times the rms noise level .
however , the significance of a resolved spectral feature is higher than its peak intensity suggests .
a better estimate of the significance of a resolved spectral feature is obtained by rebinning the spectrum . by dividing this
smoothed spectrum by its rms noise level , the peak of the spectral feature is expressed in terms of rms noise level .
the peak of the spectral feature will reach the highest significance for one specific binning .
this yields the spectral width of the feature and the significance of the integrated spectral feature .
an additional check can be made on a possibly detected maser peak .
since the spectra are averages of many individual scans , statistics can be done on this series of scans . for a specific channel
we can calculate the variance over the set of scans , yielding an estimate for the noise level of each channel .
a feature in the variance spectrum coincident with a possibly detected maser peak casts doubt on the maser nature of the peak .
a maser peak is not expected to vary intrinsically within the course of an observation , which makes an anomaly in the noise more likely .
the variance spectrum can also be used to check the derived average rms noise level . in the same way
, we can calculate the covariance between one channel and another channel over the set of scans . for each channel
, we can integrate this covariance over a fixed number of adjacent channels , obtaining the covariance spectrum .
again , a feature in the covariance spectrum right at the position of the possibly detected maser peak would suggest an anomaly in the noise instead .
we used a few different frequency settings during the course of the observation to check for systematic flux offsets due to bad channels .
the distribution of the observations over three separate shifts allows for a check on consistency too .
the very unlikely case of an artificial signal in the hrs spectrograph can be ruled out by simultaneous observation with the lrs spectrograph .
finally , proximity in velocity space to an oh detected maser feature further enhances the significance of a possible sio detection .
all hrs spectra of the sio(21)@xmath0 transition of iras045536825 are added together . the resulting spectrum is shown in fig . 1 ( a and b ) .
this spectrum is the result of a total of 9.9 hours on - source integration time . due to the beamswitching and overhead from reading / writing , tuning , and pointing checks ,
the actual observing time was @xmath6 hours .
the 1 @xmath7 level (= rms ) is only 63 mjy .
this low a noise level could only be achieved by the excellent performance of the new receiver .
the sio maser emission is clearly seen as a narrow , yet resolved peak at v@xmath8 km s@xmath4 , at a level of @xmath9 @xmath7 .
we replaced the flux value of each channel by the average flux within a bin , centered on that channel and 15 channels wide (= 2.25 km s@xmath4 ) . the spectrum is shown in fig . 1 ( c ) .
now , the significance of the maser is 6.9 @xmath7 ( 1 @xmath10 mjy ) .
the maser emission peaks at v@xmath11 km s@xmath4 , with a peak flux f@xmath12 jy ( fig . 1 ( b ) ) .
the fwhm of the feature is @xmath13 km s@xmath4 .
the total width of the peak is @xmath14 km s@xmath4 , with the center - of - gravity at v@xmath15 km s@xmath4 . integrating the feature in the region v@xmath16 km s@xmath4
, we obtain an integrated flux
f@xmath17 jy km s@xmath4 .
there is a hint of emission at 260 km s@xmath4 , as well as a possible broad pedestal emission component at v@xmath18 km s@xmath4 .
if real , this would imply an integrated flux of about a factor two higher than estimated above .
the variance and covariance spectra were featureless .
the peak is also dominantly present in the lrs spectrum that we took during 5.7 hours out of the total on - source integration time .
thus , the maser peak satisfies all the above - mentioned detection criteria .
the optical counterpart of iras045536825 is the red supergiant woh g064 ( westerlund et al .
1981 ; elias et al . 1986 ) .
it is extremely bright , both bolometric ( m@xmath19 , zijlstra et al .
1995 ) as well as in the mid - ir ( f@xmath20 jy ) .
the progenitor mass is probably as massive as @xmath21 . for this massive a star ,
the spectral type is very late : about m7 .
the severe circumstellar extinction estimated from the ( probably self - absorbed ) 9.7 @xmath22 m silicate feature seems to be inconsistent with the relatively moderate circumstellar extinction estimated from the optical .
this has usually been taken as an indication for either binarity or a highly flattened circumstellar material distribution ( roche et al .
1993 ) .
sio maser emission intensity is correlated with the mid - ir luminosity ( e.g. jewell et al .
1991 ; lane 1982 ; nyman et al .
oh / ir stars which satisfy k
l@xmath23 mag exhibit brighter sio(21)@xmath0 masers than redder stars ( nyman et al .
. the sio maser emission intensity can be highly variable in time , but typically peaks in the first quarter of the pulsation cycle ( lane 1982 ; nyman & olofsson 1986 ) . at the time of the observations
iras045536825 satisfied all of these criteria for being an outstanding candidate for detection of sio(21)@xmath0 maser emission .
iras045536825 is the first extra - galactic stellar oh maser discovered ( wood et al .
1986 ) and the strongest stellar 1612 mhz oh maser source in the lmc ; it also exhibits 1665 mhz mainline oh maser emission ( wood et al .
both oh masers are double peaked , from which wood et al . derived a stellar velocity of v@xmath24(oh)@xmath25 km s@xmath4 .
the bulk of the presently detected sio maser emission comes from a peak that is situated redward of the oh maser emission , with a velocity difference with respect to v@xmath24(oh ) of v@xmath26 km s@xmath4 .
this is atypical , since sio maser emission is usually observed to be centered at the stellar velocity ( e.g. heske 1989 ; jewell et al .
1991 ; lewis et al .
1995 ; nyman et al .
sio maser emission is known to be highly variable in both intensity and line shape in an erratic way on timescales of months ( lane 1982 ) , but excursions of sio maser peaks only reach several km s@xmath4 with respect to the stellar velocity ( e.g.bujarrabal et al .
1986 ; lane 1982 ; nyman & olofsson 1986 ; olofsson et al .
thus , it seems likely that the oh is not centred on the stellar velocity .
an indication that not all the oh emission has been observed comes from the very low value for the expansion velocity derived from the oh of only 11 km s@xmath4 : this is a factor of 3 4 lower than is normally found for supergiants . to confirm this hypothesis
, we took a medium resolution spectrum ( r @xmath27 75,000 ) with the ntt ( eso ) in october 1995 : the spectrum shows h@xmath28 emission peaking at v@xmath29 km s@xmath4 ( extending between v@xmath30 km s@xmath4 ) , confirming the revised value for the stellar velocity .
the difference between the sio maser peak velocity and the blue - most edge of the oh emission at v@xmath31 km s@xmath4 yields an expansion velocity of v@xmath32 km s@xmath4 , much more compatible with milky way rsgs than the exceptionally low outflow velocity of v@xmath33 km s@xmath4 previously determined .
the red - shifted oh emission is not observed .
the latter is expected if there is an inner ionized region which is optically thick at 18 cm ( e.g. shepherd et al .
the spectral type of iras045536825 is not consistent with an ionized stellar wind .
the discovery of forbidden [ nii ] lines ( elias et al .
1986 ) confirms the existence of ionized gas in its vicinity , but it is not clear whether this hii region is centred on the star . from the integrated flux of the maser emission , and assuming that the maser emits isotropically , we derive a total photon flux of @xmath34 photons / s .
sio maser emission is thought to be tangentially amplified ( diamond et al .
1994 ) . for maser emission coming from a rather discrete structure such as a globule
, the angle between the maser beam and the line of sight can be non - zero . in that case , the assumption of isotropic emission is no longer valid , which may lead to either over- or under - estimating the total photon flux .
assuming isotropic emission , the observed total photon flux places iras045536825 at the bright end of the distribution of the milky way agb stars .
the total photon fluxes of the rsgs vy cma and vx sgr are an order of magnitude larger ( lane 1982 ) , but they are exceptionally bright .
numerical simulations yield larger total photon fluxes with increasing stellar radius ( bujarrabal 1994 ) .
observations showed that the total photon flux also increases with increasing pulsation period ( lane 1982 ) . in this respect
, iras045536825 may not be that similar to milky way sources but rather underluminous , since from its very large radius ( it is very bright and very cool ) and its relatively long period ( p @xmath35 d ) we might have expected iras045536825 to be at least as bright as vy cma and vx sgr .
alcolea et al .
( 1990 ) found that mira variables always have a ratio sio / ir (= sio(21)@xmath0 peak intensity/8 @xmath22 m intensity ) @xmath36 ( see also hall et al .
this is the maximum pump efficiency and it is only reached in favorable conditions ( bujarrabal et al .
semi regular variables ( srvs)s with visual amplitudes exceeding 2.5 magnitude always have about the maximum efficiency , while srvs with lower amplitude in v rarely reach the maximum efficiency and are often more than an order of magnitude less efficient . vy cma and vx sgr have ratios sio / ir @xmath37 and 1/4 , and large amplitudes in v of resp . 3.1 and 7.5 mag
. this may be ( part of ) the cause why these two rsgs exhibit such bright sio maser emission .
iras045536825 has sio / ir @xmath38 , which means that it is quite efficient , but not optimal .
the small amplitude of 0.3 magnitude in k suggests that iras045536825 is not a large amplitude variable , although the amplitude in v is unknown .
iras045536825 has an integrated- over peak flux ratio of f@xmath39/f@xmath40 km s@xmath4 , whereas all detections in alcolea et al .
1990 have f@xmath39/f@xmath41 km s@xmath4 ( miras ) and f@xmath42/f@xmath43 km s@xmath4 ( srvs ) .
this is another indication that iras045536825 is probably underluminous in terms of total photon flux . in the sample of alcolea
( 1990 ) , all stars with periods exceeding 400 days have spectral types m6 or earlier .
the rsgs with detected sio(21)@xmath0 maser emission all have spectral type m4.5 or earlier , whereas detected giants have later spectral types but shorter periods ( solid symbols in fig . 2 ) .
the m4.5 rsg nml cyg ( pulsation period p @xmath44 d ) may be the milky way star with the closest resemblance to iras045536825 .
alcolea et al .
detected sio(21)@xmath0 maser emission from this star with a pump efficiency of only sio / ir @xmath45 .
however , the total photon flux ( relative to the 8 @xmath22 m intensity ) is similar to that from iras045536825 .
the same is true for their energy distributions ( elias et al .
we discovered the first extra - galactic sio maser , from the red supergiant iras045536825 in the lmc .
the sio maser peak was situated 16 km s@xmath4 redward of the center of the double peaked oh maser emission .
we argue that the sio maser peak velocity coincides with the stellar velocity .
this would mean that the outflow velocity of the circumstellar matter around iras045536825 is v@xmath46 km s@xmath4 , which is typical for galactic rsgs .
the peak intensity of the sio maser emission is not incompatible with ranges found in galactic rsgs , but the total integrated photon flux is lower than expected .
alcolea , j. , bujarrabal , v. , gmez - gonzlez , j. : 1990 , a&a 231 , 431 bujarrabal , v. , planesas , p. , gmez - gonzlez , j. , martn - pintado , j. , del romero , a. : 1986 , a&a 162 , 157 bujarrabal , v. , planesas , p. , del romero , a. : 1987 , a&a 175 , 164 bujarrabal , v. : 1994 , a&a 285 , 953 chapman , j.m . ,
cohen , r.j . : 1986 , mnras 220 , 513 diamond , p.j . , kemball , a.j . ,
junor , w. , zensus , a. , benson , j. , dhawan , v. : 1994 , apj 430 , l61 elias , j.h . , frogel , j.a . , schwering , p.b.w . : 1986 , apj 302 , 675 hall , p.j . ,
allen , d.a . ,
troup , e.r . ,
wark , r.m . ,
wright , a.e . : 1990 , mnras 243 , 480 heske , a. : 1989 , a&a 208 , 77 jewell , p.r . , snyder , l.e . , walmsley , c.m . ,
wilson , t.l . ,
gensheimer , p.d . : 1991 , a&a 242 , 211 lane , a.p . : 1982 , ph .
d. thesis , university of massachusetts lewis , b.m . ,
david , p. , le squeren , a.m. : 1995 , a&ass 111 , 237 nyman , l .-
. , johansson , l.e.b.,booth , r.s . : 1986 , a&a 160 , 352 nyman , l .- . , olofsson , h. : 1986 , a&a 158 , 67 nyman , l .- . , hall , p.j . , le bertre , t. : 1993 , a&a 280 , 551 olofsson , h. , rydbeck , o.e.h .
, nyman , l .- . : 1985 , a&a 150 , 169 reid , i.n . : 1991 , apj 382 , 143 roche , p.f . ,
aitken , d.k . ,
smith , c.h .
: 1993 , mnras 262 , 301 westerlund , b.e . ,
olander , n. , hedin , b. : 1981 , a&ass 43 , 267 spehherd , m.c .
, cohen , r.j . ,
gaylard , m.j . , west , m.e . : 1990 , nature , 344 , 522 wood , p.r . , bessell , m.s . ,
whiteoak , j.b . : 1986 , apj 306 , l81 wood , p.r . ,
whiteoak , j.b . ,
hughes , s.m.g .
, bessell , m.s . ,
gardner f.f . , hyland , a.r .
: 1992 , apj 397 , 552 zijlstra , a.a . ,
loup , c. , waters , l.b.f.m . ,
whitelock , p.a . ,
van loon , j.th . ,
guglielmo , f. : 1995 , mnras ( in press ) | we report on the detection of sio j=21 v=1 maser emission from the red supergiant iras045536825 in the lmc .
it has thereby become the first known source of sio maser emission outside the milky way . |
the search for extra - terrestrial intelligence ( seti ) has been ongoing since the publication of the seminal article by cocconi and morrison @xcite . for many years
, searches for signals were performed only at radio wavelengths despite the fact that searches at optical wavelengths had been suggested as early as 1961 @xcite , following the invention of the laser . however , spectacular improvements in laser technology over the last few decades have dramatically strengthened the case for optical searches .
we can now imagine a laser system , built with presently known technology , that could produce detectable signals over distances to nearby stars .
arguments based on signal - to noise , pulse dispersion and energy budget considerations support the idea that laser pulses may be an effective way to conduct interstellar communication @xcite . during the past decade
, the field of ` optical search for extra - terrestrial intelligence ' ( oseti ) has developed as an established sub - discipline , and several optical searches have been completed or are now under way @xcite .
here we present the results of a search for nanosecond optical transients via the solar tower atmospheric cherenkov effect experiment ( stacee ) .
the paper is organized as follows .
we begin with a description of the stacee detector and observation features relevant to the search .
we list the targets observed and the data sets obtained and we describe calibration procedures .
finally , we explain the analysis procedure and present the results of our search .
the stacee experiment was designed for high energy gamma - ray astronomy in the energy range from 100 gev to 10 tev .
it makes use of the atmospheric cherenkov technique whereby energetic astrophysical gamma rays are detected and measured via the cherenkov light generated by the relativistic particles in the air showers which result when the gamma rays interact in the upper atmosphere .
the wavelength of the light used for this technique is in the range from 300 to 600 nm .
more details on the history and present state of ground - based gamma - ray astronomy can be found elsewhere @xcite .
stacee makes use of the large steerable mirrors ( heliostats ) of the national solar thermal test facility ( nsttf ) in albuquerque , new mexico .
the nsttf , built for solar power research , has 224 heliostats , each with an area of 37 m@xmath0 , that can direct sunlight onto a tower on the south side of the field .
stacee uses 64 of these heliostats to sample the cherenkov light pool and direct the light onto secondary mirrors located on the tower .
the secondaries focus the light onto cameras , each of which comprise a cluster of photomultiplier tubes ( pmts ) such that each pmt views one heliostat .
the process is illustrated schematically in figure [ concept ] .
stacee s efficiency for detecting optical photons is wavelength - dependent . at wavelengths shorter than 350
nm the acceptance is cut off by the glass of the heliostat mirrors which are aluminized on their rear surface . at longer wavelengths , the quantum efficiency of the pmts drops to zero between 600 and 700 nm .
thus stacee s sensitivity is best between 400 and 500 nm and peaks at about 420 nm .
such a response is quite suitable for detecting cherenkov light , but it limits the range of possibilities for detection of interstellar laser pulses .
the main concern is that extinction due to absorption and scattering will be greater at shorter wavelengths ( @xmath1 at @xmath2 nm ( zagury 2001 ) ) and might render a possible signal too weak to see over a realistic distance . with the use of interstellar reddening maps
@xcite it can be shown that such an effect would amount to a loss of 10 to 20% of the beam intensity from sources in the directions of our targets .
however , it can still be argued that a longer wavelength would be preferred , and we run the risk that this would be the choice of an advanced civilization .
pulses from each pmt are amplified and then split , one copy of which is sent to trigger electronics , the other to an 8-bit 1 gigasample / s flash - analog - to - digital converter ( fadc ) .
the trigger copy is discriminated and dynamically delayed to account for the the observing geometry since arrival times of cherenkov light on each heliostat depend on the elevation and azimuth of the source being tracked .
delayed pulses from the different channels are combined and a minimum multiplicity is required within a coincidence window typically 12 ns wide .
the fadc data are written to a buffer and when a trigger is generated , a 192 ns portion of this buffer is saved to a data file .
thus the data file consists of a series of events , each of which contain 64 192-ns traces . more details of the stacee detector can be found elsewhere @xcite .
the potential of stacee as a detector that is well suited to oseti observations has been described elsewhere @xcite .
because of its large mirror area , stacee is potentially more sensitive than detectors based on conventional telescopes @xcite .
other atmospheric - cherenkov detectors have already been used in a limited way for oseti @xcite or are under consideration for future studies @xcite . while it is true that , with 64 heliostats , each with an area of 37 m@xmath0 , stacee has over 2300 m@xmath0 of collection area for oseti studies , this has to be balanced with its large field - of - view ( fov ) of about 0.6 degrees .
this is a design feature , since sensitivity to cherenkov light from air showers requires an fov of about this size .
this means that , though stacee is sensitive to low - intensity laser pulses , the air showers it detects provide a background that does not affect the optical - telescope detectors .
thus , the many pmt channels of stacee can not be used independently to increase effective area ; they must be used together to reject background from air showers generated by high - energy gamma rays and by ( the far more copious ) charged cosmic rays .
there are several criteria for rejecting backgrounds from air showers , the most powerful of which are uniformity and multiplicity .
a laser pulse from a distant source will uniformly illuminate the heliostat field , so we would expect every pmt channel in the stacee detector to report the same number of photons , though with deviations due to poisson fluctuations and channel - to - channel efficiency differences .
the data presented here were acquired between january and may 2007 . during this time , stacee was being run largely to complete a data set on the active galactic nucleus 1es 1218 + 304 @xcite . during the observing hours when this source was not near transit , we kept the detector operational and staffed in order to be ready to follow up gamma - ray burst ( grb ) alerts .
we used these stand - by hours for a series of 10-minute oseti observations on a list of candidate stars .
we selected stars from the habcat catalog @xcite , a list of nearby sun - like stars that conceivably harbor habitable planets .
the targets chosen for stacee were those that would be near zenith when they transited , since the optics of stacee are optimal for sources overhead .
all told , we observed 187 stars .
a histogram of their distances from earth is shown in figure [ distance ] .
the celestial coordinates of the stars are plotted in figure [ radec ] .
the hipparcos catalog identifier of each star and the date on which it was observed is listed in table [ targets ] .
we note that the fov of stacee is approximately 0.6 degrees so there is also sensitivity to any sources located in interstellar space in the region of the targeted stars .
( it has been argued @xcite that advanced civilizations might inhabit `` dyson spheres '' which completely enclose their host stars and are therefore not associated with visible stars .
this hypothesis was originally put forward in support of the notion of looking for extraterrestrials at infrared wavelengths .
however it does not exclude the possiblility that such beings could be sending out laser pulses of the type we can detect . ) in regions between closely - spaced stars the effective sensitivity is increased since that region of space is observed for a longer time .
this is quantified in figure [ contour ] where the points from figure [ radec ] have been smeared with a gaussian function with full - width of 0.6 degrees .
the contours represent steps of 20% of the peak sensitivity .
the data were acquired via the standard stacee trigger system , which was configured for grb follow - up observations .
we used this trigger for the oseti runs to have the capability to switch quickly to grb observations , with a minimum number of changes , following any alert from the network . for oseti running the discriminator threshold for each pmt
was set at 90 mv , which corresponds to approximately 9 photoelectrons .
almost all stacee triggers result from air showers caused by charged cosmic rays and a small number arise from gamma - ray events . in both cases ,
a pool of cherenkov light is distributed over the heliostat field and this pool has a characteristic structure in space and time .
typically there will be systematic differences from channel to channel in the amount of light received and its time of arrival .
thus a very powerful discriminant against air showers is the requirement that every pmt channel report a pulse size consistent with the arrival of a uniform flux of optical photons . to apply such a criterion
it is necessary to unfold detector effects , two of which are geometric acceptance and electronic gain . the gains of the pmts and their amplifiers are equalized with a laser calibration system @xcite but the efficiencies of the optical components ( heliostats , secondary mirrors and light concentrators ) also need to be accounted for . as described elsewhere @xcite , the effective fov of each heliostat depends on its distance from the tower , and the channel - to - channel differences are roughly compensated for by using optical concentrators with different angular acceptances attached to the pmts . additionally , collection of light from the heliostats has slight channel - dependent variations .
for example , the pmt cameras occult some of the light from the heliostats , with the exact amount depending on the particular heliostat . fortunately , we can directly measure the net result of these effects by performing drift scans on bright stars .
we point the heliostats to a location that matches the declination of a star and is four minutes ahead of it in right ascension .
we disable heliostat tracking so that each mirror is held in a fixed position .
we record photocurrents for eight minutes as the star drifts through the fov of each pmt channel .
sample plots from the first eight pmt channels are shown in figure [ varplot ] . in this figure
we plot fadc baseline variance vs time .
( the baseline variance is linearly related to photocurrent but its channel - to - channel calibration is more precisely known than that for the photocurrent . ) as the star drifts into the field - of - view , the current rises to a maximum and then returns to its quiescent value .
the curve is usually asymmetric due to details of the geometry of the heliostat position and the corresponding pmt .
we calculate the difference between the current at the four - minute mark and the average of the currents for the first and last 30 seconds of the scan and use this quantity to define the throughput for the pmt channel .
note that in some cases the maximum deviation does not occur precisely at four minutes because of a slight mis - pointing of the heliostat . for the throughput calculation , however , it is important to use the value at four minutes since that takes into account the effect of the slight mispointing , which will also be present during oseti runs .
this method of estimating net throughput is quite powerful .
since the photocurrent depends linearly on the number of photoelectrons arriving at the first dynode in the pmt , we automatically account for quantum efficiency and collection efficiency in the pmt , as well as gains in the electron multiplier structure and in the downstream amplifiers .
this is in addition to the geometric effects of light collection , reflection and occultation . after calculating a throughput for each pmt channel ,
the values are scaled such that the channel with the largest throughput has unity value .
the relative throughputs are shown in figure [ throughput ] .
most channels are within a factor of two of the most efficient one and all are within a factor of three .
the drift scans are well - suited for measuring relative response from channel to channel , but they can not be used to determine the absolute optical throughput . to do this
we use a ray - tracing program that models all the optical components using measured quantities such as reflectivities and transmissions .
an indication of the detail of the ray - tracing program is given in figure [ raytrace ] , which shows the results of tracing the trajectories of photons that arrive vertically onto the heliostat field .
the left panel shows the impact points of photons at the secondary mirror on the tower .
the outline of this mirror is shown as a circle and the blank region in the lower part is caused by occultation from the camera structure . in the right panel , impact points at the focal plane are shown ; the circle corresponds to the entrance aperture of the optical concentrator .
aberration due to coma , as well as the structure caused by the 25 facets of the heliostat can be seen . using the ray - tracing program
we calculate an average effective area per heliostat of approximately 3 m@xmath0 at a wavelength of 420 nm .
the effective areas range from 1.5 m@xmath0 to 3.5 m@xmath0 . at 1.5 m@xmath0 a relatively large amount of light is lost due to occultation and the mis - match between the heliostat image size and the optical concentrator acceptance ; at 3.5 m@xmath0 , such losses are less severe .
as noted above , the nominal area of each heliostat is 37 m@xmath0 , so the overall efficiency is less than 10% .
most of this is due to the quantum efficiency of the pmts . to calculate the optical fluxes to
which stacee would be sensitive , we used a simple monte carlo program to simulate the response of our detector to uniform light pulses .
fluxes ( in photons / m@xmath0 ) were simulated over a range of values , and the mean numbers of photons expected in each pmt channel were calculated via the optical throughputs obtained from the drift scans .
the pmt channel with the best throughput was assigned an effective area of 3.5 m@xmath0 , and the other channels were scaled accordingly .
the resulting mean values for the expected numbers of detected photons were used to generate a characteristic poisson distribution for each pmt channel .
64-channel events were generated by sampling these distributions and the number of detected photons in each channel was corrected for the relative acceptance of the channel .
thus the mean corrected photoelectron count should be the same for all channels but the fluctuations depend on the relative acceptances of the pmt channels .
an event was accepted if all channels had a corrected photoelectron count above threshold .
results of this calculation are shown in figure [ accept_curve ] where efficiency as a function of flux level is plotted for four different threshold values . for typical thresholds ,
the detector is sensitive to fluxes of the order of 10 - 15 photons / m@xmath0 .
this is very much a wavelength - dependent statement ; these calculations were carried out assuming a laser wavelength of 420 nm . for longer wavelengths
our sensitivity is less , due to the fall - off in quantum efficiency of the pmts .
each data run resulted in a file that contained approximately 1500 events , each of which comprised 64 fadc traces , each 192 samples long .
the fadcs used an 8-bit digitizer so that each sample corresponded to a number from 0 to 255 . a typical fadc trace consisted of a fluctuating baseline , with an average value of about 220 counts , with a negative - going pulse between sample numbers 80 and 120 .
ancillary data such as the time of each event , as read from a gps clock , made up the rest of the data file .
the data analysis program was designed to re - apply the trigger criterion with tighter timing windows and select events with acceptance - corrected pulses which were consistent , allowing for poisson fluctuations , with coming from a uniform flash of light .
the data processing was done in four stages . in the first stage ,
the first 50 fadc samples for each pmt channel were histogrammed and fit to a gaussian distribution .
the mean was saved as the pedestal for the channel and the root - mean - square standard deviation was saved to be used in defining a cut threshold .
the early part of the trace was used to avoid biases from in - time pulses .
after the pedestal was computed , it was subtracted and the trace was inverted .
( see see figure [ sample_data ] for an example of typical traces . ) any channel with at least one sample in its trace above a cut of 3 standard deviations ( typically 15 - 20 counts ) was deemed to be hit .
if , in the raw trace , any samples were at zero ( i.e. , the pulse had `` bottomed out '' ) the channel was labelled as saturated ( indicating a light level beyond the dynamic range of the digitizer ) . in the second stage ,
hit but unsaturated pmt channels were used to calculate an event time .
channels with maximum sample size more than 50 counts above baseline were used for timing , with the time of the maximum sample used as the time of the pulse .
the average of these times from all available channels was used as the event time ; due to details of the trigger , this fluctuated by order 10 ns from event to event .
the deviation from event time for each channel was histogrammed and fit , at the end of the run , to a gaussian to define a timing window ( mean plus or minus 2.5 standard deviations ) for use in the following stage .
the width of the timing window so defined was channel - dependent with most channels having a width of 10 ns , though some at the far edge of the heliostat field had windows that were 25 ns wide . the use of narrow , channel - specific search windows enabled the use of the lowest possible thresholds in looking for valid pulses . in the third stage , the event time and the window limits were used to define a search region for each pmt channel .
the maximum sample within the window was used to define the peak of the pulse and the pulse charge was computed by summing the 15 samples starting 5 samples before this point .
the channel was counted as hit if the maximum sample was above the cut value defined above .
( note that a channel defined as hit in the second stage of the analysis could lose this status at this point because the above - threshold sample in stage 2 was from outside the search window . )
this procedure produced a charge for all channels , even those that were below trigger threshold .
the analysis can be summarized while viewing figure [ sample_data ] .
displayed in this figure are traces from eight fadc channels for an arbitrary event .
the central portion of each trace ( samples 61 through 156 ) is displayed in the figure .
the upper line below each pulse delineates the 15 ns integration window used for charge estimation . the lower line indicates the range over
which samples are searched for a pulse . for events with at least 55 pmt channels that have a sample above threshold in the search window , all 64 integrated charges were saved in a file for further processing .
although the signature of a laser pulse would be 64 hit channels , the cut at 55 allows for dead channels in some runs , as well as poisson fluctuations on low - intensity candidates . for the selected events ,
charges were converted from units of counts - ns to photoelectrons with a scale factor of 20 counts - ns per photoelectron .
this factor was determined via special calibration and gain - equalizaton runs made with the laser calibration system @xcite .
gaussian ( @xmath3 ) uncertainties , which are statistically adequate for @xmath4 , were assigned to the photoelectron estimates at this point .
the photoelectron estimates and their uncertainties were multiplied by the relative throughput corrections ( figure [ throughput ] ) . using only pmt channels that reported a corrected photoelectron count of at least 10 , we computed an average number of photoelectrons for the event .
the threshold was applied to avoid including channels that , for some reason were not functioning properly during the run .
( on some occasions one or two heliostats were stowed due to mechanical problems . )
if an event passed further selection cuts , the low - charge channels could be included in the calculations after checking on the hardware status .
typically , more than 50 channels contributed to the average photoelectron calculation . to search for events consistent with uniform light illumination ,
we characterize the extent to which the measured flux in all pmt channels is statistically consistent with a single value .
we define a standard chi - squared statistic to characterize the extent to which channels deviate from the the average value .
the chi - square per degree - of - freedom ( @xmath5 ) for such a large number of statistically independent channels should be very strongly peaked at a value of 1.0 for events that are consistent with uniform illumination .
events with ( @xmath5 ) significantly greater than 1.0 are considered to be non - uniform , and are tagged as air showers .
for example , for n=50 the distribution expected for laser pulses is approximately normal with a standard deviation of 0.3 .
events with @xmath6 were individually checked .
an example of data from such an event is shown in figure [ passed - ex - pe ] .
the corrected number of photoelectrons for each of the 64 pmt channels is plotted vs channel number , along with the fitted average .
figure [ passed - ex - means ] shows the mean charge ( in counts - ns ) vs channel number for all events in the run .
the uneven structure is caused by the distribution of cherenkov light in air showers and the efficiency for triggering on such showers as a function of where they hit the heliostat field .
the fact that no channels reported zero for mean charge indicates that they were all in good working order for this run .
this result indicates that the zeros in pmt channels 45 - 48 in figure [ passed - ex - pe ] are due to a lack of light in the corresponding heliostats .
this is not unexpected in air - shower events .
the distribution of @xmath5 for all fits in the run is shown in figure [ passed - ex - chi2 ] .
most runs have one or two events that have @xmath6 at this stage .
these are invariably rejected because they have pmt channels that have below - threshold pulse sizes ( often zero ) even though the mean - charge distributions indicate that they are in good working order .
these small pulses were left out of the initial average and @xmath5 calculation because of the 10-photoelectron threshold designed to accomodate possibly bad channels .
when they are included , the @xmath5 gets much larger . as can be seen in figure [ accept_curve ] , with the 10-photoelectron threshold used in the analysis , we were fully efficient for flashes that resulted in flux levels above 10 photons / m@xmath0 at the detector .
after processing all the data runs , we did not have any candidates consistent with the arrival of pulses of at least this intensity .
in the field of elementary particle physics , searches for new particles and related phenomena are common .
examples include the hunt for the weak interaction bosons ( w and z ) , the top quark and neutrino oscillations .
these searches were ultimately successful . presently ongoing searches include the quest for axions , the higgs boson , and dark matter particles such as wimps . in such searches
there is a space of possibilities with coordinates such as particle mass and interaction cross - section .
null results are presented as exclusion plots , which indicate what parts of the parameter space have been ruled out in that they do not warrant further consideration .
progress is made by shifting the boundaries of these excluded regions into unexplored regions .
oseti searches are not so systematic .
it is possible to imagine repeating the same experiment some time after it was first carried out and obtaining different results .
a null result could be changed to a positive result in the event a distant civilization has , in the interim started sending out detectable signals .
repeating the experiment with a slight change in one aspect , e.g. , dwell time on the target , might lead to a successful detection .
what did we learn from the results of our search with the stacee detector ?
we can say that blue - green laser pulses of detectable intensity were not being sent , in our direction and from the vicinity of the target star , within those 10-minute intervals during which our search was directed at them .
saying more is difficult .
instrumentally , we have learned that use of a heliostat array for oseti searches is possible .
however the large mirror area is reduced to a much smaller effective area by the quantum efficiency of the pmts and various geometric factors .
the wavelength range is rather small and this reduces the amount of parameter space that can be investigated , though this shortcoming could be reduced with pmts chosen for oseti rather than cherenkov astronomy . the large number of heliostats can not be used to synthesize a giant collector , and , thereby , increase sensitivity to faint laser pulses .
this is because multiple channels are needed to reject the huge background of nanosecond optical flashes caused by extensive air showers .
the simple analysis described in this paper requires a larger photon flux than the 2 photons / m@xmath0 foreseen by covault ( 2001 ) but it is not excluded that , using techniques described in that paper ( fitting the times to a wavefront , cutting on arrival direction , _ etc . _ ) , one could lower the required flux .
the loss of effective collection area is offset somewhat by the larger field - of - view afforded by an instrument of this type .
we have not explored how few channels can be used to suppress cherenkov background sufficiently , but it could be possible to divide the detector into several smaller detectors to allow for several regions to be monitored at once this would require more electronics to construct the extra triggers . how does the stacee search compare to other oseti efforts ?
the harvard - princeton search @xcite was much more extensive , with nearly 16,000 observations on over 6000 objects totalling 2400 hours of data .
the typical time per target was 24 minutes .
the stated sensitivity was 100 photons / m@xmath0 in a window of 5 ns in the 450 - 650 nm band for the harvard telescope .
the princeton telescope , which observed in parallel for 1142 of the targets , had a similar sensitivity ( 80 photons / m@xmath0 ) .
a team at the lick observatory made 5000 targeted observations with use of a conventional optical telescope @xcite .
their flux sensitivity was similar , but their time - on - target was 10 minutes .
the harvard all - sky - detector team @xcite reported that they achieved a sensitivity of 95 photons / m@xmath0 in a window of 3 ns .
they ran in a drift - scan mode with a consequent one - minute dwell time per source point .
a group @xcite searching for oseti signals in archival data from the whipple 10-meter gamma - ray telescope , another cherenkov - based instrument , demonstrated that a sensitivity of 10 photons / m@xmath0 is possible .
each run in their search had a length of 28 minutes , but the same part of the sky was targeted in many different runs since gamma - ray observations typically require very long exposures . because of this the total time spent on a given target was very large .
one conclusion from this comparison is that the large mirror areas of the cherenkov telescopes can be used to improve flux sensitivity by about one order of magnitude compared with the smaller optical telescopes . to put the flux of 10 photons / m@xmath0 in perspective
, we consider a 400 nm laser at a distance of 1000 light years .
a 0.3 mj pulse would contain @xmath7 photons , and if these were formed into a diffraction - limited beam with a 10 m mirror , the beam diameter at the earth would be @xmath8 m. assuming that the profile is such that half the photons fall within a diameter of @xmath9 m , an average flux density of 10 photons / m@xmath0 is obtained . lasers with an order of magnitude greater pulse energy have already been developed , so our search is not limited by what might be considered reasonable technical achievements of distant civilisations .
the stacee detector was dismantled during the summer of 2007 so there will be no further observations of the type reported here . some members of the stacee collaboration ,
however , have been studying the possibilities for future detectors dedicated to oseti .
one idea that has emerged is the concept of multiple , distributed oseti detectors .
this idea foresees the use of a large number of inexpensive detectors , each of which carry out drift scans in the manner of the harvard all - sky detector @xcite .
a detector element would consist of two pairs of telescopes , where each telescope comprises a simple spherical mirror , approximately one meter in diameter , with a single pmt at the focus .
the two members of a pair , both pointing at the zenith , would be located on the order of 100 meters apart and operated in coincidence mode with cables and simple electronics . with each local coincidence
, digitized pulses from the two pmts would be stored along with a gps time stamp .
most of these recorded events would be due to cherenkov light from large air - showers , but the rate would not be large due to the separation of the two members of the pair . a second , identical pair directed at the same spot on the sky would be deployed several kilometers away such that coincidences between the two pairs could not be caused by air - showers .
data from the two pairs would be compared off - line using the gps time stamps to identify coincidences and the digitized pulses to check for uniform illumination .
a large number of such installations could be built , possibly in the style of the various outreach projects involving cosmic - ray detectors located at high schools @xcite . with a large number of detectors , each scanning the same strip of sky night after night
, one would address what we consider to be one of the most important limitations of the oseti searches carried out to date : duty factor .
it is important to be live and on - target when a distant laser is pointed our way .
sensitivity to faint laser pulses would not be useful if the laser and detector were not co - aligned at the crucial time .
a world - wide grid with spacing on the order of the detector field - of - view would be useful and affordable , given the simple nature of the detectors , which involve very basic electronics and no moving parts .
of course , the success of an oseti campaign would depend on whether civilizations exist that have the impetus to transmit laser pulses .
the number of such civilizations can not be estimated reliably , though limits can be set by using tools such as the drake equation @xcite .
this equation attempts to identify the various factors needed to make such an estimate , so that some of the uncertainties involved can be quantified .
hetesi and regly @xcite purported that the drake equation is not complete in that it neglects certain key factors , e.g. , advanced civilisations may choose to remain silent .
there are many good reasons to suspect that such civilisations would be in the majority .
we have used the stacee high - energy gamma - ray detector to look for fast blue - green laser pulses from the vicinity of 187 stars .
these stars have been chosen because their characteristics are such that they may harbor habitable planets and they are relatively close to earth .
each star was observed for 10 minutes , and we found no evidence for laser pulses with intensities greater than about 10 photons / m@xmath0 at 420 nm in any of the data sets .
we are grateful to the staff at the national thermal solar test facility for their enthusiastic and professional support .
the stacee project was funded in part by the u.s .
national science foundation , the university of california , los angeles , the natural sciences and engineering research council , le fonds qubecois de la recherche sur la nature et les technologies , the research corporation , and the california space institute .
armada , a. , cortina , j. and martinez , m. ( 2004 ) optical seti with magic , in _ neutrinos and explosive events in the universe _ , edited by maurice m. shapiro , todor stanev and john p. wefel , springer pp 307 - 310 .
bhathal , r. ( 2001 ) optical seti in australia , in _ proceedings spie 4273 : the search for extraterrestrial intelligence ( seti ) in the optical spectrum iii _ , edited by s.a .
kingsley and r. bhathal , spie , bellingham , wa , pp 144 - 152 .
carlson , b.e . ,
brobeck , e. , jillings , c.j . ,
larson , m.b . ,
lynn , t.w . ,
mckeown , r.d . , hill , james e. , falkowski , b. , seki , r. , sepikas , j. , yodh , g.b .
( 2005 ) search for correlated high energy cosmic ray events with chicos _ journal of physics _
g31 , pp 409 - 416 .
chantell m.c . , bhattacharya d. , covault c.e . , dragovan m. , fernholz r. , gregorich d.t . , hanna d.s . ,
marion g.h .
, ong r.a . , oser s. , tumer t.o . , williams d.a .
( 1998 ) prototype test results of the solar tower atmospheric cherenkov effect experiment ( stacee ) , _ nuclear instruments and methods _
a408 pp 468 - 485 .
covault , c.e .
( 2001 ) large area solar power heliostat arrays for oseti , in _ proceedings spie 4273 : the search for extraterrestrial intelligence ( seti ) in the optical spectrum iii _ , edited by s.a .
kingsley and r. bhathal , spie , bellingham , wa , pp 161 - 172 .
gingrich , d.m .
, boone , l.m . , bramel , d. , carson , j. , covault , c.e . ,
fortin , p. , hanna , d.s . ,
hinton , j.a . , jarvis , a. , kildea , j. , lindner , t. , mueller , c. , mukherjee , r. , ong , r.a . ,
ragan , k. , scalzo , r.a . ,
theoret , c.g . ,
williams , d.a . , zweerink , j.a.,(2005 ) , the stacee ground based gamma - ray detector , _ ieee transactions on nuclear science _ 52 : pp 2977 - 2985 .
hanna , d.s . , bhattacharya , d. , boone , l.m . ,
chantell , m.c . , conner , z. , covault , c.e . ,
dragovan , m. , fortin , p . , gregorich , d.t . ,
hinton , j.a . ,
mukherjee , r. , ong , r.a .
, oser , s. , ragan , k. , scalzo , r.a . ,
schuette , d.r . ,
theoret , c.e . ,
tumer , t.o . ,
williams , d.a .
, zweerink j.a.,(2002 ) , the stacee-32 ground based gamma - ray detector , _ nuclear instruments and methods _
a489 : pp 126 - 151 .
holder , j. , ashworth , p. , lebohec , s. , rose , h.j . , weekes , t. , ( 2005 ) optical seti with imaging cherenkov telescopes , in _ proceedings of the 29th international cosmic ray conference , pune _ 5 , pp 387 - 390 .
howard , a.w . ,
horowitz , p. , wilkinson , d.t . ,
coldwell , c.m .
, groth , e.j . , jarosik , n. , latham , d.w . ,
stefanik , r.p . ,
willman , a.j.jr . ,
wolff , j. , zajac , j.m . , ( 2004 ) search for nanosecond optical pulses from nearby solar - type stars , _ astrophysical journal _ 613 , pp 1270 - 1284 .
howard , a.w . , horowitz , mead , c. , sreetharan , p. , gallicchio , j. , howard , s. , coldwell , c. , zajac , j. , sliski , a. , ( 2007 ) , initial resulsts from the harvard all - sky optical seti , _ acta astronautica _ 61 , pp 78 - 87 .
mukherjee , r. , akhter , n. , ball , j. , carson , j.e .
, covault , c.e . ,
driscoll , d.d . ,
fortin , p. , gingrich , d.m . ,
hanna , d.s . , jarvis , a. , kildea , j. , lindner , t. , mueller , c. , ong , r.a . ,
ragan , k. , williams , d.a . ,
zweerink , j.a . , ( 2007 ) , stacee observations of 1es 1218 + 304 , in _ proceedings of the 30th international cosmic ray conference , merida_. weekes , t.c .
( 2006 ) , revealing the dark tev sky : the atmospheric cherenkov imaging technique for very high energy gamma - ray astronomy , in _ proceedings of the international workshop on `` energy budget in the high energy universe '' _ , kashiwa , japan .
an illustration of the concept behind the stacee gamma - ray detector .
an incident gamma ray produces an air shower and the particles therein generate cherenkov light .
heliostats on the ground direct this light towards a tower .
secondary mirrors on the tower focus this light onto nearby photomultiplier tubes .
each pmt views a unique heliostat.,scaledwidth=100.0% ] declination ( in degrees ) vs right ascension ( in hours ) for the stars observed during the stacee oseti campaign .
the space of possibilities was limited by the requirement that the stars transit close to zenith.,scaledwidth=100.0% ] contour plot of sensitivity as a function of target direction .
the positions of the stars plotted in figure [ radec ] have been smeared with a gaussian function having a full - width corresponding to the stacee field - of - view ( approximately 0.6 degrees ) .
the contour step size is 20% of the peak sensitivity.,scaledwidth=100.0% ] fadc baseline variances ( in digital - counts - squared ) vs time ( in minutes ) for a drift scan with a bright star .
each plot corresponds to a different pmt channel .
the deviation of the trace at 4 minutes from the baseline value is a relative measure of the net throughput of the channel.,scaledwidth=100.0% ] results for pmt channel 49 from the ray - tracing program used to evaluate absolute throughput .
left panel : impact points of photons at the plane of the secondary mirror .
the 2-m circle indicates the position of the mirror .
the blank area at the lower right is caused by occultation from the camera structure .
right panel : photon impact positions at the entrance to the optical concentrator on the pmt .
comatic aberration as well as structure due to the heliostat facets can be seen.,scaledwidth=100.0% ] efficiency vs optical flux levels for the stacee detector for four threshold values .
an effective area of 3.5 m@xmath0 for the most efficient pmt channel has been assumed and all other channels have been scaled according to the relative throughputs plotted in figure [ throughput ] . using a 10-photoelectron analysis threshold , we find no candidate laser flashes in the stacee data .
, scaledwidth=100.0% ] sample traces from eight pmt channels of a particular event .
the channel numbers are shown below each trace .
samples 61 through 156 ( of a possible 192 saved samples ) are displayed .
the upper line shows the range over which the pulse is integrated to give an estimate of the charge . the lower line below indicates the window used in searching for a valid pulse . although the window for the pulse search can vary from channel to channel , the integration gate is fixed at 15 ns .
note that the flat tops of the traces for channels 34 and 35 are due to saturation.,scaledwidth=100.0% ] example of an event singled out for attention during the final stage of analysis .
the corrected pulse size ( in photoelectrons ) is plotted vs pmt channel number , together with the average computed using channels reporting more than 10 photoelectrons .
the @xmath5 for this event is 1.64.,scaledwidth=70.0% ] mean charge vs pmt channel number for the run containing the candidate event shown in figure [ passed - ex - pe ] .
units are counts - ns . the structure is due to array geometry , trigger effects , and the distribution of light from air showers . the fact that each channel reports a non - zero mean value indicates that there were no defective channels in this run.,scaledwidth=70.0% ] l l l l l l l l 41998 & 0124 & 36053 & 0217 & 47993 & 0222 & 76859 & 0317 + 42686 & 0124 & 37374 & 0217 & 49770 & 0222 & 78751 & 0317 + 41930 & 0124 & 38995 & 0217 & 71753 & 0223 & 79538 & 0317 + 45334 & 0124 & 39112 & 0217 & 72892 & 0223 & 80206 & 0317 + 47313 & 0124 & 39459 & 0217 & 72175 & 0223 & 81147 & 0317 + 46033 & 0125 & 41530 & 0217 & 74260 & 0223 & 82432 & 0317 + 47016 & 0125 & 42445 & 0217 & 44502 & 0312 & 36952 & 0318 + 43596 & 0125 & 42763 & 0217 & 45883 & 0312 & 37991 & 0318 + 43895 & 0125 & 42405 & 0217 & 46707 & 0312 & 38950 & 0318 + 45126 & 0125 & 45175 & 0217 & 48202 & 0312 & 41382 & 0318 + 45307 & 0125 & 46446 & 0217 & 48465 & 0312 & 41888 & 0318 + 32304 & 0213 & 25614 & 0218 & 49283 & 0312 & 43024 & 0318 + 34809 & 0213 & 26863 & 0218 & 50920 & 0312 & 44502 & 0318 + 34977 & 0213 & 27361 & 0218 & 49241 & 0312 & 45053 & 0318 + 36904 & 0213 & 28102 & 0218 & 51454 & 0312 & 45006 & 0318 + 38480 & 0213 & 29617 & 0218 & 52885 & 0312 & 46993 & 0318 + 26506 & 0216 & 30391 & 0218 & 44646 & 0314 & 72628 & 0319 + 28005 & 0216 & 32168 & 0218 & 48566 & 0314 & 73829 & 0319 + 27234 & 0216 & 32239 & 0218 & 49157 & 0314 & 74512 & 0319 + 28707 & 0216 & 32211 & 0218 & 51804 & 0314 & 75579 & 0319 + 29319 & 0216 & 33557 & 0218 & 52601 & 0314 & 76525 & 0319 + 29874 & 0216 & 35193 & 0218 & 52119 & 0314 & 77402 & 0319 + 31619 & 0216 & 34812 & 0218 & 41382 & 0316 & 78638 & 0319 + 31431 & 0216 & 35825 & 0218 & 41642 & 0316 & 80326 & 0319 + 34468 & 0216 & 38295 & 0218 & 43471 & 0316 & 72760 & 0320 + 35962 & 0216 & 38949 & 0218 & 45006 & 0316 & 74156 & 0320 + 37620 & 0216 & 39667 & 0218 & 45162 & 0316 & 75908 & 0320 + 37589 & 0216 & 41292 & 0218 & 46993 & 0316 & 76656 & 0320 + 39031 & 0216 & 42386 & 0218 & 48387 & 0316 & 78277 & 0320 + 40037 & 0216 & 41270 & 0218 & 49529 & 0316 & 78559 & 0320 + 41272 & 0216 & 42730 & 0218 & 49687 & 0316 & 80807 & 0320 + 41521 & 0216 & 44525 & 0218 & 49111 & 0316 & 77469 & 0416 + 42104 & 0216 & 45484 & 0218 & 50767 & 0316 & 77670 & 0416 + 41691 & 0216 & 46314 & 0218 & 52052 & 0316 & 79381 & 0416 + 44987 & 0216 & 47413 & 0218 & 77304 & 0316 & 80571 & 0416 + 44559 & 0216 & 40398 & 0221 & 79105 & 0316 & 80128 & 0416 + 24295 & 0217 & 41454 & 0221 & 79824 & 0316 & 81677 & 0416 + 25030 & 0217 & 41888 & 0221 & 80417 & 0316 & 83075 & 0416 + 25852 & 0217 & 42182 & 0221 & 81776 & 0316 & 84283 & 0416 + 26849 & 0217 & 43024 & 0221 & 82938 & 0316 & 83326 & 0416 + 27501 & 0217 & 44856 & 0221 & 45196 & 0317 & 84688 & 0416 + 28727 & 0217 & 45053 & 0221 & 47739 & 0317 & 85879 & 0416 + 30751 & 0217 & 47581 & 0221 & 48081 & 0317 & 87526 & 0416 + 31972 & 0217 & 47191 & 0221 & 50185 & 0317 & 70297 & 0420 + 31576 & 0217 & 46636 & 0222 & 49358 & 0317 & 70930 &
0420 + 32658 & 0217 & 46022 & 0222 & 51460 & 0317 & 87680 & 0420 + 32568 & 0217 & 47575 & 0222 & 51687 & 0317 + | we have used the stacee high - energy gamma - ray detector to look for fast blue - green laser pulses from the vicinity of 187 stars .
the stacee detector offers unprecedented light - collecting capability for the detection of nanosecond pulses from such lasers .
we estimate stacee s sensitivity to be approximately 10 photons / m@xmath0 at a wavelength of 420 nm .
the stars have been chosen because their characteristics are such that they may harbor habitable planets and they are relatively close to earth .
each star was observed for 10 minutes and we found no evidence for laser pulses in any of the data sets . 0.5 cm |
previous measurements of black hole ( bh ) x - ray transients in quiescence show that the bh systems have x - ray luminosities ranging from @xmath6 to @xmath7 erg s@xmath3 @xcite . as in outburst
, there is evidence for the presence of an accretion disk in quiescence @xcite , but the accretion rate onto the bh itself is unclear as is the nature of the accretion flow . although there is currently no consensus on the origin of the quiescent x - ray emission , three emission sites are considered to be viable for at least some systems : the accretion disk ; the putative outflow or jet ; and the secondary star .
one picture for the quiescent accretion flow is that of an inner quasi - spherical and optically thin region surrounded by a optically thick disk as in the advection - dominated accretion flow ( adaf ) model @xcite .
this model explains the faintness of quiescent bhs as being due to advection of radiation across the bh event horizon @xcite .
however , it has been determined that adafs are convectively unstable for certain viscosity ranges , leading to convection - dominated accretion flows ( cdaf , quataert & gruzinov 2000 ) .
such a model can explain the low bh luminosities without the need for advection across an event horizon .
in addition , bh advection may not be necessary if outflows or jets transport the accretion energy out of the system @xcite .
there is evidence for the presence of compact radio jets from bh systems at low x - ray luminosities @xcite , and these may persist in quiescence .
such jets may possibly contribute to the x - ray emission @xcite .
finally , @xcite suggested the secondary as an x - ray emission site , but we note that this possibility has been ruled out for most systems @xcite . in this work ,
we describe results of observations of the bh transients v4641 sgr and xte j1859 + 226 in quiescence made using the _ chandra x - ray observatory _ @xcite .
these systems both had major x - ray outbursts in 1999 @xcite .
while the xte j1859 + 226 outburst was typical of bh transients , the behavior of v4641 sgr was unusual with a long period of low level activity punctuated by a 12 crab flare lasting 1 - 2 days . from optical observations covering the 67.6 hr v4641 sgr orbit and the ( likely ) 9.16 hr xte j1859 + 226 orbit ,
the compact object masses are above 3 , indicating that these systems contain bhs @xcite . at the time of the outburst
, v4641 sgr exhibited a one - sided , rapidly evolving , relativistic radio jet @xcite , and there is evidence that xte j1859 + 226 also produced a radio jet @xcite .
the main motivations of our _ chandra _ observations are to study the bhs in quiescence and to search for large - scale x - ray jets like those detected for xte j1550564 @xcite and 4u 1755338 @xcite . here
, we report on the detection of the bhs in quiescence , and we will report on the x - ray jet search and the consequences of the non - detection of those jets in a future paper .
we obtained 2 _ chandra _ observations of v4641 sgr in 2002 : a 4.3 ks observation on ut august 5 ( obs .
1a ) ; and a 25.3 ks observation on ut october 21 ( obs .
for both observations , we used the advanced ccd imaging spectrometer ( acis ) in imaging mode , and the target was placed on one of the back - illuminated acis chips ( s3 ) .
observation 1a was carried out under director s discretionary time and was prompted by x - ray , optical , and radio activity from the source during 2002 may and june @xcite .
considering the possibility that the source might still be active in x - rays during observation 1a , we used a 1/4 ccd subarray to mitigate photon pileup effects .
we found that the source was not bright @xcite , and we used the full ccd for observation 1b .
we also obtained a 24.8 ks _
observation of xte j1859 + 226 on 2003 february 5 ( obs .
2 ) using acis in imaging mode .
the last known x - ray activity from xte j1859 + 226 was in 2000 @xcite .
for all 3 observations , the background levels remain low throughout , allowing us to use the full exposure time .
vla ( very large array ) radio observations occurred within 1 week of observations 1a and 1b and within 2 weeks of observation 2 .
v4641 sgr and xte j1859 + 226 showed no radio emission with rms noise levels near 0.1 mjy / beam at 4 and 6 cm .
we produced 0.3 - 8 kev acis images using the `` level 2 '' event lists from the standard data processing .
for the analysis described below , we used the ciao ( _ chandra interactive analysis of observations _ ) v2.3 software and caldb ( calibration data base ) v2.21 . to obtain the maximum sensitivity for source detection in the v4641 sgr field , we combined the data for observations 1a and 1b .
we restricted our search for sources to a 2 arcmin@xmath2 region containing the position of v4641 sgr .
this region , which is shown in figure [ fig : v4641 ] , is fully covered by both observations giving a spatially uniform exposure time of 29.5 ks .
we used the ciao routine `` wavdetect '' with a detection threshold of @xmath8 @xcite to search for sources in the combined image , leading to the detection of 12 sources with between 6 and 36 counts per source . a source with 9 counts ( labeled 1 in figure [ fig : v4641 ] )
is detected at a position consistent with the v4641 sgr radio position . as the v4641 sgr radio position is known to @xmath9 @xcite
, it is worthwhile to obtain the best possible x - ray position in order to test whether _ chandra _ source 1 is , in fact , v4641 sgr . to register the image ,
we cross - correlated the _
chandra _ source positions with the 2 micron all - sky survey ( 2mass ) ir sources in the field .
we determined that 4 of the 12 _ chandra _ sources have 2mass sources within the _ chandra _ pointing uncertainty of @xmath10 .
for these 4 sources , the angular separations between the 2mass and _ chandra _ positions range from @xmath11 to @xmath12 . given the surface density of 2mass sources ( @xmath13 sources arcsec@xmath14 down to k@xmath15 ) , there is a 0.14% probability that a match with the largest separation is spurious .
as source 1 ( the v4641 sgr candidate ) is identified as one of the 2mass sources ( 2mass 181921632524258 with ir magnitudes j = @xmath16 , h = @xmath17 , and k@xmath18 = @xmath19 ) , we used the other 3 sources to register the image . for these 3 sources , the average 2mass
to _ chandra _ differences in r.a . and decl .
are @xmath20 and @xmath21 arcseconds , respectively , where the uncertainties account for the _ chandra _ statistical position error as well as the @xmath22 position uncertainties for 2mass .
we performed the indicated shifts ( @xmath23 east and @xmath9 south ) to complete the registration of the _ chandra _ image to the 2mass positions .
the position of source 1 is r.a .
= 18h 19 m 21s.641 , decl . = 25@xmath24 24@xmath25 25@xmath26 ( equinox 2000.0 , systematic uncertainty = @xmath27 , statistical uncertainty = @xmath27 ) .
the best _ chandra _ position is @xmath28 from the 2mass position and @xmath29 from the radio position obtained from a vla image obtained on 1999 september 17 , for which @xcite interpreted the emission as coming from the black hole rather than the extended jet . after registering the image
, we find that all 9 acis counts are contained within @xmath30 of the radio position , while the expected number of background counts in a @xmath30 circle is 0.28 .
thus , the poisson probability that the detection is spurious is @xmath31 .
the n and e lines are @xmath32 .
the 12 detected sources are circled , and sources 1 - 4 ( marked with 2 circles ) have 2mass identifications .
source 1 is v4641 sgr .
[ fig : v4641],scaledwidth=40.0% ] we carried out a similar analysis for xte j1859 + 226 .
we searched for _ chandra _ sources in a 3 arcmin@xmath2 region centered on the xte j1859 + 226 radio position @xcite using wavdetect .
as shown in figure [ fig:1859 ] , we detected 5 _ chandra _ sources with between 6 ( source 1 ) and 113 ( source 2 ) counts .
source 1 is consistent with the xte j1859 + 226 radio position .
we cross - correlated the _ chandra _ positions with the 2mass positions and found that sources 3 - 5 have likely ir counterparts .
for these 3 sources , the maximum separation between the _ chandra _ and 2mass positions is @xmath33 , and there is a 0.07% probability that a match with this separation is spurious . using the 2mass identifications , we determined the position shifts required to register the _ chandra _ image , and these are @xmath34 and @xmath35 arcseconds in r.a . and decl .
, respectively .
after shifting the _ chandra _ image by @xmath36 west and @xmath12 south , the position of source 1 is r.a .
= 18h 58 m 41s.485 , decl .
= + 22@xmath24 39@xmath25 29@xmath37 ( equinox 2000.0 , systematic uncertainty is @xmath38 , statistical uncertainty is @xmath39 in r.a . and
@xmath40 in decl . ) , which is consistent with the @xmath41 vla radio position @xcite .
all 6 acis counts are contained within @xmath30 of the radio position , while the expected number of background counts in a @xmath30 circle is 0.18 .
thus , the poisson probability that the detection is spurious is @xmath42 .
although the number of counts detected for both v4641 sgr and xte j1859 + 226 is small , we extracted energy spectra ( with 9 and 6 counts , respectively ) . for v4641 sgr
, the photon energies range from 0.87 kev to 7.0 kev , and the mean photon energy is 3.5 kev . for xte j1859 + 226
, the photon energies range from 0.75 kev to 2.7 kev , and the mean photon energy is 1.5 kev .
we fitted the energy spectra with a power - law model with interstellar absorption .
we used this model because this spectral shape is typically seen for quiescent bh systems @xcite , but we emphasize that we can not rule out other spectral shapes .
we fitted the spectra using cash statistics @xcite , and we fixed the column density ( @xmath43 ) to the h i values from @xcite : @xmath44 cm@xmath14 and @xmath46 cm@xmath14 for v4641 sgr and xte j1859 + 226 , respectively .
we determined the errors on the parameters by producing and fitting 10,000 simulated spectra .
we used the best fit parameters from the fits to the actual data as input to the simulations .
the power - law photon indices are @xmath47 for v4641 sgr and @xmath48 for xte j1859 + 226 ( 90% confidence errors ) .
the energy spectra of quiescent bhs are typically well - described by an absorbed power - law with a photon index of @xmath49 - 2.3 @xcite .
although a ( much ) better v4641 sgr spectrum should be obtained to confirm its spectral shape , the best estimate of @xmath50 is well outside of the typical range , and even at the upper limit of @xmath51 , v4641 sgr would be the hardest of the quiescent bhs .
however , the uncertain column density is an important source of systematic error , and we re - fitted the v4641 sgr spectrum leaving @xmath43 as a free parameter .
our simulations indicate 90% confidence upper limits on @xmath52 and @xmath43 of 2.5 and @xmath53 cm@xmath14 , respectively .
thus , we conclude that either the v4641 sgr is intrinsically hard or that @xmath43 is higher than the interstellar value . .
the n and e lines are @xmath32 .
the 5 detected sources are circled , and and the 3 sources marked with 2 circles have 2mass identifications .
source 1 is xte j1859 + 226.[fig:1859],scaledwidth=37.0% ] we used the spectral fits to obtain flux and luminosity measurements for the sources . for v4641 sgr , the absorbed 0.3 - 8 kev flux is @xmath54 erg cm@xmath14 s@xmath3 ( 90% confidence errors ) , and the unabsorbed luminosity in the same energy band is @xmath0 ( @xmath1/7 kpc)@xmath2 erg s@xmath3 , where @xmath1 is the source distance .
optical observations of v4641 sgr in quiescence give a distance of @xmath55 kpc ( orosz et al .
, in prep . ) . for xte j1859 + 226 ,
the absorbed 0.3 - 8 kev flux is @xmath56 erg cm@xmath14 s@xmath3 , and the luminosity is @xmath57 ( @xmath1/11 kpc)@xmath2 erg s@xmath3 . the distance estimate of 11 kpc comes from the x - ray and optical properties of the source @xcite .
as confirmed black hole systems ( based on compact object mass measurements ) , v4641 sgr and xte j1859 + 226 represent important additions to the group of bhs that have previously been detected in x - rays in quiescence . of the 15 confirmed @xcite transient bh systems , 14 now have quiescent x - ray detections or sensitive upper limits , and only grs 1915 + 105 , which has been in outburst since its discovery in 1992 , does not .
the x - ray luminosity ( @xmath58 ) measurements are summarized in figure [ fig : luminosity ] .
we include a new low x - ray flux measurement for xte j1550564 @xcite , that has not been previously considered in the context of studies of quiescent bhs , and we assume a distance of 5.3 kpc @xcite to convert from flux to luminosity .
we also include the lowest flux measurements for the recurrent transient gx 3394 @xcite , and we assume a distance of 4 kpc .
the median luminosity is @xmath5 erg s@xmath3 for the 9 systems with previous detections , and the measured luminosities for v4641 sgr and xte j1859 + 226 are consistent with the median . for quiescent bhs ,
the x - ray emission is thought to originate from the accretion flow , a jet , or the secondary star , and , here , we discuss how our results for v4641 sgr may constrain theoretical models for x - ray production . the v4641 sgr source distance ( 7 kpc ) along with the proper motion of the one - sided radio jet seen in 1999 ( @xmath59 per day ) indicate that the angle between the jet axis and our line of sight is @xmath608@xmath24 @xcite .
thus , if the x - ray emission originates in a fast - moving jet , one expects v4641 sgr to be brighter due to relativistic beaming .
however , the v4641 sgr x - ray luminosity ( assuming isotropic emission ) is similar to the luminosities of the other bh systems , for which the jet axes are either not known or are known to be relatively far from our line of sight , indicating that the x - ray emission is not highly beamed and limiting the velocity of a putative x - ray emitting jet . for a continuous synchrotron x - ray jet @xcite , it is unlikely that the bulk - motion lorentz factor could be higher than @xmath611.5 as this would cause the source to be brighter than an unbeamed source by a factor of @xmath6110 for a jet that is @xmath62 from our line of sight @xcite .
this calculation assumes a spectrum with a photon index of 1.5 , which is in the range of values that can be produced within the @xcite model . assuming a photon index of 0.2 ( our best estimate for v4641 sgr ) would lead to a somewhat higher limit on the lorentz factor , but it is unclear whether such a hard spectrum could have a synchrotron origin .
also , we note that the lorentz factor constraint is similar for models where x - rays are produced in jets via inverse comptonization @xcite .
the v4641 sgr constraint on the lorentz factor ( @xmath631.5 ) is consistent with recent results for bh systems in the canonical low - hard state @xcite . for accretion disk models such as the adaf model
, the quiescent x - ray luminosity depends mainly on mass accretion rate rather than on system orientation .
thus , the luminosities of v4641 sgr and xte j1859 + 226 would be expected to be similar to the other bh systems , as observed .
if the hard v4641 sgr x - ray spectrum is confirmed and is intrinsic to the source , it would suggest that another parameter besides accretion rate is important .
hard spectra can be produced by adafs @xcite and cdafs @xcite , but it is unclear why one system at close to the median luminosity would be intrinsically much harder than the others .
the spectral type of the v4641 sgr secondary is b9 iii @xcite , making it the most luminous of the bh transients . in quiescence , the ratio of the x - ray to bolometric luminosity ( @xmath64 ) for v4641 sgr is @xmath61@xmath65 based on the x - ray luminosity reported here and the optical luminosity of the secondary ( orosz et al . , in prep . ) . the v4641 sgr ratio is higher than the average value but comparable to the highest values measured by _
rosat _ for late b - type stars @xcite .
while it is unlikely that the hard x - ray flux from v4641 sgr comes from the secondary , we can not rule out the possibility that the secondary makes a some contribution to the soft x - ray flux below @xmath612 kev
. the early type secondary could also be important if it has a strong wind .
the wind could collide with accretion disk material , leading to additional x - ray production , or it could cause extra x - ray absorption , explaining the hard v4641 sgr spectrum .
we would like to thank h. tananbaum for granting director s discretionary time and the referee for very useful comments .
the two micron all sky survey is a joint project of the univ . of massachusetts and ipac / caltech , funded by nasa and nsf .
the national radio astronomy observatory is a facility of the nsf operated under cooperative agreement by associated universities , inc .
jat acknowledges partial support from _ chandra _ award number go3 - 4040x .
pk acknowledges partial support from nasa grant nag5 - 7405 and _ chandra _ award number go3 - 4043x .
jmm thanks the nsf . | using the _ chandra x - ray observatory _ , we have detected the black hole transients v4641 sgr and xte j1859 + 226 in their low luminosity , quiescent states .
the 0.3 - 8 kev luminosities are @xmath0 ( @xmath1/7 kpc)@xmath2 erg s@xmath3 and @xmath4 ( @xmath1/11 kpc)@xmath2 erg s@xmath3 for v4641 sgr and xte j1859 + 226 , respectively . with the addition of these 2 systems , 14 out of the 15 transients with confirmed black holes ( via compact object mass measurements ) now have measured quiescent luminosities or sensitive upper limits . the only exception is grs 1915 + 105 , which has not been in quiescence since its discovery in 1992 .
the luminosities for v4641 sgr and xte j1859 + 226 are consistent with the median luminosity of @xmath5 erg s@xmath3 for the systems with previous detections .
our analysis suggests that the quiescent x - ray spectrum of v4641 sgr is harder than for the other systems in this group , but , due to the low statistical quality of the spectrum , it is not clear if v4641 sgr is intrinsically hard or if the column density is higher than the interstellar value . focusing on v4641 sgr
, we compare our results to theoretical models for x - ray emission from black holes in quiescence .
also , we obtain precise x - ray positions for v4641 sgr and xte j1859 + 226 via cross - correlation of the x - ray sources detected near our targets with ir sources in the 2 micron all - sky survey catalog . |
with the increasing precision of the measurements of the cosmic microwave background ( cmb ) anisotropies and large scale structures ( lss ) of the universe as well as various other astronomical observations , it is now possible to have a clear and consistent picture of the history and content of the universe since nucleosynthesis . although the now widely used term of `` standard model of cosmology '' might remain premature as our knowledge of the cosmological scenario is by far less precise than that of the standard model of particle physics , the matter content of the universe as well as its expansion rate are now known within a precision of a few percents with great confidence .
it is also well established that the cosmological perturbations which gave rise to the cmb anisotropies and the lss of the universe was inflationary like , with a close to scale invariant harrison - zeldovitch spectrum . moreover , the measurement of both the temperature and polarization anisotropies of the cosmic microwave background allows to test the paradigm of adiabaticity of the cosmological perturbations and hence the precise nature of the mechanism whih has generated them .
the simplest realizations of the inflationary paradigm predict an approximately scale invariant spectrum of adiabatic and gaussian curvature fluctuations , whose amplitude remains constant outside the horizon , and therefore allows cosmologists to probe the physics of inflation through observations of the cmb anisotropies and the lss matter distribution . however , this is certainly not the only possibility .
models of inflation with more than one field generically predict that , together with this so - called adiabatic component , there should also be entropy , or isocurvature perturbations @xcite , associated with fluctuations in number density between different components of the plasma before photon decoupling , with a possible statistical correlation between the adiabatic and isocurvature modes @xcite .
baryon and cold dark matter ( cdm ) isocurvature perturbations were proposed long ago @xcite as an alternative to adiabatic perturbations .
a few years ago , two other modes , neutrino isocurvature density and velocity perturbations , have been added to the list @xcite .
moreover , it is well known that entropy perturbations seed curvature perturbations outside the horizon @xcite , so that it is possible that a significant component of the observed adiabatic mode could be strongly correlated with an isocurvature mode .
such models are generically called _ curvaton models _ @xcite , and are now widely studied as an alternative to the standard paradigm . furthermore
, isocurvature modes typically induce non - gaussian signatures in the spectrum of primordial perturbations . in this paper , we describe in more detail the analysis performed in ref @xcite , constraining the various isocurvature components .
we also extend it by including additional observational constraints , and extra free parameters in the model .
we use data from the temperature power spectrum and temperature - polarization cross - correlation measured by the wmap satellite @xcite ; as well as from the small - scale temperature anisotropy probed by vsa @xcite , cbi @xcite and acbar @xcite ; from the matter power spectrum measured by the 2-degree - field galaxy redshift survey ( 2dfgrs ) @xcite and the sloan digital sky survey ( sdss ) @xcite ; and also use data from the recent type ia supernova compilation of ref .
we do not use the data from lyman-@xmath2 forests , since they are based on non - linear simulations carried under the assumption of adiabaticity .
the first bounds on isocurvature perturbations assumed uncorrelated modes @xcite , but recently also correlated ones were considered in refs .
our general analysis includes this possibility . in the first part of this work , we will not assume any specific model of inflation , nor any particular mechanism to generate the perturbations ( late decays , phase transitions , cosmic defects , etc . ) , and thus will allow all five modes adiabatic ( ad ) , baryon isocurvature ( bi ) , cdm isocurvature ( cdi ) , neutrino isocurvature density ( nid ) and neutrino isocurvature velocity ( niv ) to be correlated ( or not ) with each other , and to have arbitrary tilts . in terms of model building ,
the simplest situation beyond the paradigm of adiabaticity is that of a single isocurvature mode mixed with the adiabatic one .
therefore , we shall not consider more than one isocurvature mode at a time , and our primordial perturbations will be described by three amplitudes and three spectral indices , associated respectively with the adiabatic , isocurvature and cross - correlated components .
this choice is somewhat different from that of refs .
@xcite , who introduce several modes at a time , but a single tilt for each power spectrum of primordial perturbations .
the assumption that all the modes have comparable amplitudes and a common tilt are both difficult to motivate theoretically and , to our knowledge , all proposed mechanisms based on inflation stand far from this case . for instance , double inflation leads to at most one isocurvature mode , always with a tilt differing from the adiabatic one ; on the other hand , the curvaton scenario predicts a single tilt , but only one isocurvature mode , fully correlated or anti - correlated with the adiabatic mode .
however , in the absence of any theoretical prior , we believe the approach of refs .
@xcite is interesting and complementary to ours . for simplicity
, we will neglect the possible effects of spatial curvature and tensor perturbations , and assume that neutrinos are massless @xcite .
each model will be described by eleven cosmological parameters : the six usual parameters of the standard @xmath3cdm model ; the amplitude and spectral index of the primordial isocurvature perturbation ; the amplitude and spectral index of the cross - correlation angle between the adiabatic and isocurvature modes ; and finally , a parameter @xmath4 describing the equation of state of dark energy , assumed to be time - independent as a first approximation . in addition , we will treat conservatively the matter - to - light bias of the 2df and sdss redshift surveys as two extra free parameters . in section
ii we describe the notation we use in our analysis of isocurvature modes and discuss the relation with multi - field inflation , and in particular two - field models , on which we will concentrate ourselves . in section iii
we discuss the general bounds on our full parameter space from cmb , lss and sn data using a bayesian likelihood analysis .
in section iv we analyze those bounds in a concrete model of two - field inflation : double inflation , with two uncoupled massive fields .
a particular case is that in which the two fields have equal masses , like in complex field inflation , which we show is not ruled out . in section
v we present our general conclusions .
for the theoretical analysis , we will use the notation and some of the approximations of refs . @xcite . during inflation
more than one scalar field could evolve sufficiently slowly that their quantum fluctuations perturb the metric on scales larger than the hubble scale during inflation .
these perturbations will later give rise to one adiabatic mode and several isocurvature modes .
we will restrict ourselves here to the situation where there are only two fields , @xmath5 and @xmath6 , and thus only one isocurvature and one adiabatic mode . introducing more fields
would complicate the inflationary model and even then , it would be rather unlikely that more than one isocurvature mode contributes to the observed cosmological perturbations .
the evolution during inflation will draw a trajectory in field space .
perturbations along the trajectory ( i.e. in the number of @xmath7-folds @xmath8 ) will give rise to curvature perturbations on comoving hypersurfaces , @xmath9 while perturbations orthogonal to the trajectory will give rise to gauge invariant entropy ( isocurvature ) perturbations , @xmath10 for instance , entropy perturbations in cold dark matter during the radiation era can be computed as @xmath11 . in order to relate these perturbations during the radiation era with those produced during inflation
, one has to follow the evolution across reheating .
during inflation we can always perform an instantaneous rotation along the field trajectory and relate the gauge - independent perturbations in the fields @xcite , @xmath12 , with perturbations along and orthogonal to the trajectory , @xmath13 and @xmath14 , @xmath15 \delta s\end{array}\right ) = \left(\begin{array}{rr}\cos\varphi & \sin\varphi\\[2 mm ] -\sin\varphi & \cos\varphi\end{array}\right ) \left(\begin{array}{c}\delta\hat{\phi}_1\\[2 mm ] \delta\hat{\phi}_2\end{array}\right)\,,\ ] ] with @xmath16 the rotation angle between the two frames . in this case
, the curvature and entropy perturbations on super - horizon scales can be written , in the slow - roll approximation , as @xcite @xmath17 & & \,{{\cal s}}_k = \frac{2}{3}\dot{\varphi}\,{\dot\phi_1\delta\hat{\phi}_2- \dot\phi_2\delta\hat{\phi}_1\over\dot\phi_1 ^ 2+\dot\phi_2 ^ 2 } = \frac{2}{3}\dot{\varphi}\,{\delta s_k\over\dot\sigma } \ , .
\end{aligned}\ ] ] where @xmath18 .
the problem however is that , contrary to the case of purely adiabatic perturbations , the amplitudes of both curvature and entropy perturbations do not remain constant on superhorizon scales , but evolve with time @xcite . in particular , due to the conservation of the energy momentum tensor , the entropy perturbations seed the curvature perturbations , and thus their amplitude during the radiation dominated era evolves according to @xcite @xmath19 & & \,\dot{{\cal s}}_k = \beta(t)h\,{{\cal s}}_k\,,\end{aligned}\ ] ] where @xmath2 and @xmath20 are time - dependent functions characterizing the evolution during inflation and radiation eras .
a formal solution can be found in terms of a transfer matrix , relating the amplitude at horizon crossing during inflation ( * ) with that at a later time during radiation , @xmath21 { { \cal s}}\end{array}\right ) = \left(\begin{array}{lr } 1 & t_{{{\cal r}}{{\cal s}}}\\[2 mm ] 0 & t_{{{\cal s}}{{\cal s}}}\end{array}\right ) \left(\begin{array}{c } { { \cal r } } _ * \\[2 mm ] { { \cal s } } _ * \end{array}\right)\,,\ ] ] where the transfer functions are given by @xcite @xmath22\ , , \\[2 mm ] & & t_{{{\cal r}}{{\cal s}}}(t , t _ * ) = \int_{t_*}^t\alpha(t')h(t')t_{{{\cal s}}{{\cal s}}}(t',t_*)dt'\,.\end{aligned}\ ] ] note that in the absence of primordial isocurvature perturbation , @xmath23 , the curvature perturbation remains constant and no isocurvature perturbation is generated during the evolution .
this is the reason for the entries @xmath24 and @xmath25 , respectively , in the transfer matrix .
note also that in many models , @xmath26 and @xmath27 vanish after reheating , so that @xmath28 remain constant during radiation domination on super - horizon scales .
however , this is not true for instance when the fluid carrying the isocurvature perturbations has a significant background density ( compared to the total universe density ) , as assumed in the curvaton scenario .
since @xmath5 and @xmath6 are essentially massless during inflation , we can treat them as free fields whose fluctuations at horizon crossing have an amplitude @xmath29 , where @xmath30 is the rate of expansion at the time the perturbation crossed the horizon @xmath31 , and @xmath32 are gaussian random fields with zero mean , @xmath33 and @xmath34 .
now , since @xmath35 and @xmath36 and just rotations of the field fluctuations , they are also gaussian random fields of amplitude @xmath37 . however , the time evolution ( [ evolution ] ) will mix those modes and will generically induce correlations and non gaussianities .
therefore , the two - point correlation function or power spectra of both adiabatic and isocurvature perturbations , as well as their cross - correlation , can be parametrized with three power laws , i.e. three amplitudes and three spectral indices , @xmath38 \delta_{{\cal s}}^2(k)\!&\equiv&\ ! { k^3\over2\pi^2}\langle{{\cal s}}^2\rangle = b^2\,\left({k\over k_0}\right)^{{n_{\rm iso}}-1}\,,\\[2 mm ] \delta_{{{\cal r}}{{\cal s}}}^2(k)\!&\equiv&\!{k^3\over2\pi^2}\nonumber \langle{{\cal r}}{{\cal s}}\rangle \ , \nonumber\\[2 mm ] & = & a\,b\,\cos\delta_{k_0 }
\left({k\over k_0}\right)^{{n_{\rm cor}}+{{1\over2}}({n_{\rm ad}}+{n_{\rm iso}})-1}\,,\nonumber\end{aligned}\ ] ] where @xmath39 is some pivot scale . since the time of horizon crossing @xmath40 in the transfer functions ( [ tf ] ) is scale - dependent ,
the correlation angle @xmath41 is in general a function of @xmath42 . in the above definitions
, we approximated @xmath43 by a power law with amplitude @xmath44 and tilt @xmath45 .
so , we assumed implicitly that the inequality @xmath46 holds over all relevant scales . in the following analysis
, we will impose that for each value of @xmath44 the tilt @xmath45 is restricted to the interval in which the inequality holds between @xmath47mpc@xmath1 and @xmath48mpc@xmath1 , which is roughly the range probed by our cmb and lss data sets . far from that range
, we expect that next - order terms become important and that the power - law approximation breaks down .
the fact that @xmath45 can be non - zero was already considered in a recent paper @xcite .
the angular power spectrum of temperature and polarization anisotropies seen in the cmb today can be obtained from the radiation transfer functions for adiabatic and isocurvature perturbations , @xmath49 and @xmath50 , computed from the initial conditions @xmath51 and @xmath52 , respectively , and convolved with the initial power spectra , @xmath53 ^ 2\ , \left({k\over k_0}\right)^{{n_{\rm ad}}-1}\,,\\[2 mm ] c_l^{\rm iso}\!&\equiv&\,\int { dk\over k}\ , \left[\theta_l^{\rm iso}(k)\right]^2\ , \left({k\over k_0}\right)^{{n_{\rm iso}}-1}\,,\nonumber\\[2 mm ] c_l^{\rm cor}\!&\equiv&\,\int { dk\over k}\ , \theta_l^{\rm
ad}(k)\,\theta_l^{\rm iso}(k)\ , \left({k\over k_0}\right)^{{n_{\rm cor}}+{{1\over2}}({n_{\rm ad}}+{n_{\rm iso}})-1}\,,\nonumber\end{aligned}\ ] ] then , the total angular power spectrum reads @xmath54 in many works ( see for instance @xcite ) , the following parametrization is employed : @xmath55 where the global normalization has been marginalized over , and @xmath56 is the entropy to curvature perturbation ratio during the radiation era , @xmath57 . we will use here a slightly different notation , used before by other groups @xcite , where @xmath58 and @xmath59 ( up to a global normalization factor ) , so that @xmath60 the parameter @xmath2 runs from purely adiabatic ( @xmath61 ) to purely isocurvature ( @xmath62 ) , while @xmath20 defines the correlation coefficient , with @xmath63 corresponding to maximally correlated(anticorrelated ) modes .
there is an obvious relation between both parametrizations : @xmath64 this notation has the advantage that the full parameter space of @xmath65 is contained within a circle of radius @xmath66 .
the north and south rims correspond to fully correlated ( @xmath67 ) and fully anticorrelated ( @xmath68 ) perturbations , with the equator corresponding to uncorrelated perturbations ( @xmath69 ) .
the east and west correspond to purely isocurvature and purely adiabatic perturbations , respectively .
any other point within the circle is an arbitrary admixture of adiabatic and isocurvature modes .
note that interpreting the results in non fully correlated and anti correlated models is made difficult by the fact that as soon as @xmath70 , the contours one obtains depend on the choice of the pivot scale .
for example in the simple case where @xmath71 , the relative amplitude between @xmath72 and @xmath73 is unchanged when one changes the pivot scale , and hence @xmath2 is unchanged , but the correlation angle @xmath74 , and hence @xmath20 changes with the pivot scale as @xmath75 . in our example
, if @xmath76 , points within the @xmath65 circle are shifted vertically toward the edge of the circle when one increases the pivot scale and shifted toward the horizontal @xmath77 line when one decreases the pivot scale .
we should emphasize that the three amplitude parameters @xmath78 $ ] , @xmath2 and @xmath20 are defined at @xmath79 , and that comparing bounds from various papers is straighforward only when the pivot scale is the same . throughout this paper
, we will use @xmath80 mpc@xmath1 , which differs from our previous work @xcite , but has the advantage of matching most of the literature .
this value of @xmath42 corresponds roughly to multipole number @xmath81 .
we computed the bayesian likelihood of each cosmological parameter with a monte carlo markov chain method , using the public code cosmomc @xcite with option mpi in the cosmos supercomputer . for each case
, we ran 32 markhov chains , and obtained between 30,000 and 60,000 samples ( without including the multiplicity of each point ) .
we used the public code camb @xcite in order to compute the theoretical prediction for the @xmath82 coefficients of the temperature and polarization power spectra , as well as the matter spectra @xmath83 , for all four different components ( the bi mode is just a rescaling of the cdi mode , see below ) .
moreover , we slightly modified the interface between camb and cosmomc in order to include the cross - correlated power spectra , as well as three independent tilts @xmath84 in the cosmomc jargon , the three tilts and the three amplitudes were implemented into the code as `` fast parameters '' , in order to save a considerable amount of time .
the likelihood of each model was then computed using the detailed information provided on the experimental websites ( or directly from the corresponding routines of the cosmomc code when available ) , using three groups of data sets : i ) cmb data : 1398 points from wmap ( @xmath85 ) @xcite , 8 points from vsa ( @xmath86 ) @xcite , 8 points from cbi ( @xmath87 ) @xcite , 7 points from acbar ( @xmath88 ) @xcite ; ii ) lss data : 32 points from 2dfgrs ( up to @xmath89 mpc@xmath1 ) @xcite , 17 points from sdss ( up to @xmath90 mpc@xmath1 ) @xcite ; and iii ) snia data : 157 points from riess et al .
we have checked that the inclusion of the hst @xcite and bbn priors are irrelevant for the determination of the hubble parameter and the baryon density ; that is , the data from cmb , lss and snia is enough in order to determine these parameters , and therefore we ignored the priors in the final analysis .
+ + + + + + + + + in order to take into account the @xmath20-dependent constraint on @xmath91 , see eq .
( [ ncorrconst ] ) , we choose to define a new parameter @xmath92 whose boundaries are fixed once and for all by the values of the pivot scale and the scales ( @xmath93 , @xmath94 ) introduced in the previous section .
the basis of cosmological parameters used by the markov chain algorithms consists on : ( i ) the seven parameters describing the standard adiabatic @xmath3cdm model , extended to dark energy with a constant equation of state , and ( ii ) the four parameters describing the admixture of one ( correlated ) isocurvature mode , already defined in the previous section .
more explicitely , we use the following basis : 1 ) the overall normalisation parameter @xmath78 $ ] where @xmath95 is the curvature perturbation in the radiation era ; 2 ) the adiabatic tilt @xmath96 ; 3 ) the baryon density @xmath97 ; 4 ) the cold dark matter density @xmath98 ; 5 ) the ratio @xmath99 of the sound horizon to the angular diameter distance multiplied by 100 ; 6 ) the optical depth to reionization @xmath100 ; 7 ) the dark energy equation of state parameter @xmath4 ; then , for the isocurvature sector ; 8) the isocurvature contribution @xmath2 ; 9 ) the cross - correlation parameter @xmath20 ; 10 ) the isocurvature tilt @xmath101 ; 11 ) the parameter @xmath102 which determines the cross - correlation tilt @xmath45 ; and finally , two arbitrary bias parameters associated to the 2df and sdss power spectrum .
our full parameter space is therefore 13-dimensional .
we did not devote a specific analysis to the case of the baryon isocurvature modes , which is qualitatively similar to that of cdi modes , since the spectra are simply rescaled by a factor @xmath103 ( @xmath104 for the cross - correlation ) : thus , compared to the ad + cdi case , significantly larger values of @xmath2 will be allowed in the ad + bi case .
like in other recent analyses , we find that the inclusion of isocurvature modes does not improve significantly the goodness - of - fit of the cosmological model , since in the ad+cdi , ad+nid and ad+niv cases the minimum @xmath105 is always between 1672 and 1674 for 1614 degrees of freedom , to be compared with 1674 for 1618 degrees of freedom in the pure adiabatic case .
therefore , the question is just to study how much departure from the standard picture is allowed , by computing the bayesian confidence limit on the isocurvature parameters .
a more detailed analysis of model comparison with bayesian information criteria @xcite will be done in a follow up paper . on fig .
[ 1dplots ] we plot the marginalised likelihood for our basis of eleven cosmological parameters , in the cases ad+cdi , ad+nid and ad+niv , compared with the pure adiabatic case .
figure [ 1dplots_bis ] shows the likelihood of some derived parameters .
it appears that most parameters are robust against the inclusion of isocurvature perturbations ( this is in agreement with the conclusion of ref .
@xcite that with only one isocurvature mode present , no significant parameter degeneracy pops out ) . our 95% c.l . on @xmath2 in the three cases
is given in table [ alpha_bounds ] .
@xmath106^{1/2}$ ] ad+cdi & @xmath107 & @xmath108 to @xmath109 ad+nid & @xmath110 & @xmath111 to @xmath112 ad+niv & @xmath113 & @xmath114 to @xmath115 [ alpha_bounds ] note that in the limit @xmath61 , the three parameters @xmath20 , @xmath116 , @xmath102 become irrelevant .
so , the fact that pure adiabatic models are very good fits implies that these parameters are loosely constrained .
this explains why the corresponding likelihoods on fig . [ 1dplots ] are not well - peaked like for other parameters .
in addition , these likelihoods should be considered with great care , because it is difficult for the markhov chains to explore in detail the tails of the multi - dimensionnal likelihood corresponding to tiny values of @xmath2 , where basically any value of ( @xmath20 , @xmath116 , @xmath102 ) are allowed .
therefore , increasing the number of samples would tend to flatten these likelihoods , while the other ones would remain stable ( as we checked explicitely ) .
however , it is clear that all models prefer a large isocurvature tilt and saturate the bound @xmath117 that we fixed in the present analysis .
this feature is important for understanding our results and comparing with other analyses , as explained in the last paragraph of this section . on fig .
[ fig : alpha ] , we plot the two - dimensional confidence levels directly for the isocurvature and cross - correlation coefficients ( @xmath2 , @xmath118 ) in the three cases ad+cdi , ad+nid and ad+niv
. the last plot corresponds to the ad+cdi case with a prior @xmath119 which is relevant for the bounds on double inflation , but the results are not substantially different from the general ad+cdi case .
we see that the ad+cdi model slightly prefers anti - correlated cases ( note that @xmath120 means a positive contribution from the cross - correlated component ) , while ad+nid and ad+niv models clearly prefer correlated ones . on figure [ fig : spectra ] , we plot the cmb and lss power spectra for two particular cdi and nid models . in order to get a better understanding of our bounds
, we chose models with large values of @xmath2 , still allowed at the 2-@xmath121 level : respectively , @xmath122 and @xmath123 .
the detailed values of other cosmological parameters for these models are given in the figure caption . in the cdi example
, one can see that the non - adiabatic contributions to all spectra remain tiny , excepted for the matter power spectrum on scales @xmath124mpc@xmath1 , due to the large isocurvature tilt @xmath125 of the model .
this is an indication that our @xmath2 bound in the cdi case depends very much on constraints on the small - scale matter power spectrum , while future improvements in the determination of cmb spectra would not reduce it dramatically .
note that a precise experimental determination of the amplitude parameter @xmath126 ( which is mainly sensitive to scales around @xmath127 mpc@xmath1 ) would probably not change things either , since our cdi models have roughly the same @xmath126 values as purely adiabatic well - fitting models ( see figure [ 1dplots_bis ] ) ; on the other hand , any constraint on smaller wavelengths could improve the bounds .
this means in particular that our choice not to include the lyman-@xmath2 data plays a crucial role in our results .
the same conclusions apply to the nid model of fig .
[ fig : spectra ] .
in addition , in the nid case , we see that the non - adiabatic contribution is significant also for small - scale ( @xmath128 ) temperature and polarization spectra . indeed
, it is well - known that the nid isocurvature and cross - correlated modes can mimick the adiabatic cmb spectra to a better extent than cdi modes ( essentially because the amplitude of the secondary peaks is not strongly supressed ) .
so , in the nid case , future improvement in the cmb data should help to improve the bounds on the isocurvature fraction .
our bounds are difficult to compare with those of ref .
@xcite , because of our different parameter space , observational data set and conventions of normalization . in the ad+cdi case ,
our analysis is closer to the one of the wmap team @xcite , although we have more free cosmological parameters ( @xmath91 , @xmath4 ) , more data ( sdss , cbi , vsa ) and less constraints on the matter to light bias .
the 95% limit @xmath129 obtained by wmap would correspond to @xmath130 in our notations , which is significantly smaller than our results .
also , the wmap @xmath131 bounds on @xmath101 are @xmath132 , while we find that the likelihood peaks at our maximum allowed value @xmath133 .
the most likely explanation is that the use of the lyman-@xmath2 data in the wmap analysis eliminates all our well - fitting models with @xmath134 and large @xmath2 values .
similar conclusions apply to our previous results @xcite , in which we did not use any lyman-@xmath2 information , but adopted a flat prior @xmath135 ( this was the interval in which our grid of models was computed ) . then , most of the well - fitting models of @xcite had slighly negative values of @xmath136 .
therefore , translating our previous results in terms of a pivot scale @xmath137mpc@xmath1 would lead to a small decrease in the @xmath2 bounds , making them comparable with the wmap bound in the cdi case , and smaller than the conservative bounds of this work .
in this section we will use the previous bounds to constrain the parameters of a concrete model of inflation called double inflation @xcite .
this is an inflationary model with two massive fields , @xmath5 and @xmath6 , of different masses , with ratio @xmath138
. no coupling ( except gravitational ) between fields is assumed .
the equations of motion and friedmann equation can be written as @xmath139 & & \ddot\phi_2 + 3h\,\dot\phi_2 + m_2 ^ 2 \phi_2 = 0\,,\\[2 mm ] & & h^2 = { \kappa^2\over6}\big(\dot\phi_1 ^ 2 + \dot\phi_2 ^ 2 + m_1 ^ 2 \phi_1 ^ 2 + m_2 ^ 2 \phi_2 ^ 2\big)\,,\end{aligned}\ ] ] where @xmath140 . in the slow - roll approximation , @xmath141
, we have the solutions @xcite @xmath142 with @xmath143 the number of @xmath7-folds to the end of inflation . for simplicity , we restrict ourselves to the case where @xmath5 and @xmath6 remain positive during inflation . substituting into the rate of expansion , @xmath144 and , using the equations of motion
, one can integrate out @xmath145 to obtain @xmath146 as inflation proceeds , @xmath147 decreases and @xmath99 also decreases .
we will call @xmath148 to the number of @xmath7-folds before the end of inflation when the scale corresponding to our hubble radius today exited during inflation .
the great advantage of double inflation is that it is possible to find explicit formulas for the perturbations on superhorizon scales .
the growing mode solutions for the scalar fields and the scalar metric perturbation in the longitudinal gauge , and in the slow - roll approximation , are given by @xcite @xmath149 & & { \delta\phi_1\over\dot\phi_1 } = { c_1(k)\over h } + c_3(k ) { 2h\phi_2 ^ 2\over r^2\phi_1 ^ 2 + \phi_2 ^ 2}\,,\\[2 mm ] & & { \delta\phi_2\over\dot\phi_2 } = { c_1(k)\over h } - c_3(k ) { 2hr^2\phi_1 ^ 2\over r^2\phi_1 ^ 2 + \phi_2 ^ 2}\,.\end{aligned}\ ] ] since @xmath5 and @xmath6 are independent uncoupled scalar fields and essentially massless during inflation , we can use the general formalism of section ii , and write , in the slow - roll approximation , @xmath150 & & c_3(k ) = - { 3h_k\over2\sqrt{2k^3}}\left({e_1({\bf k})\over m _ 1 ^ 2\phi_1 } - { e_2({\bf k})\over m_2 ^ 2\phi_2}\right)\,,\end{aligned}\ ] ] where @xmath30 is the rate of expansion when the perturbation of wavenumber @xmath151 left the horizon during inflation , where the scale of our present horizon is @xmath152 mpc .
we will now assume that the light scalar field decays at the end of inflation into the ordinary particles , giving rise to photons , neutrinos , electrons and baryons , while the cold dark matter ( cdm ) arises from the decay of the heavy field . in principle , part of the cdm
could also be produced by the light field or the heavy field could also decay into ordinary particles , but we will ignore these possibilities here . then , the perturbations in the comoving gauge take the form @xmath153 and there is only one isocurvature mode , the cdi mode , @xmath154 all of which are gauge invariant quantities . during the radiation era
, the initial conditions of all these modes are described in terms of only two @xmath42-dependent quantites , @xmath155 and @xmath156 .
the pure adiabatic initial conditions are given by the gravitational potential during the radiation era , @xmath157 with @xmath158 the amplitude of the growing adiabatic mode during inflation .
on the other hand , the isocurvature initial conditions in the radiation era arise from the perturbations in the heavy field @xmath5 at the end of inflation . in the long wavelength limit ,
the perturbations of this field during reheating follow closely the field itself , so that its energy density perturbations satisfy , in the comoving gauge , @xmath159 which is constant during inflation , across reheating and into the radiation era .
the entropy perturbation is dominated by the cdm density perturbation during the radiation era , @xmath160 . using the values of @xmath158 and @xmath161 during inflation ,
we can finally write @xmath162 & & { { \cal s}_{\rm rad}}(k ) = { \kappa h_k\over\sqrt{2k^3}}{1\over\sqrt{s_k } } \left({e_1({\bf k})\over \sin\theta_k } - { r^2\,e_2({\bf k})\over \cos\theta_k}\right)\,.\end{aligned}\ ] ] note that the cross - correlation amplitude @xmath43 is almost independent of @xmath42 , due to the cancellation of the factor @xmath163 between @xmath164 and @xmath165 , and to the fact that @xmath166 is a mild function of @xmath42 , see eq . ( [ thetas ] ) . in this case
, the tilt @xmath45 vanishes .
the adiabatic and isocurvature tilts can be computed explicitly using ( [ ks ] ) @xmath167 & & { n_{\rm ad}}= 1 - { \partial\ln{{\cal p}_{\cal r}}(k)\over\partial\ln s}\ , , \hspace{2 mm } { n_{\rm iso}}= 1 - { \partial\ln{{\cal p}_{\cal s}}(k)\over\partial\ln s}\ , , \hspace{5mm}\end{aligned}\ ] ] as functions of the angle @xmath99 and the mass ratio @xmath168 , @xmath169 { n_{\rm iso}}\!&=\ ! & 1 - { ( r^2 - 1)(r^6\tan^4\theta-1)\over s(1+(r^2 - 1)\sin^2\theta)}\times\\[2mm]\label{niso } & & \hspace{1cm}{1\over(1+(r^2+r^4)\tan^2\theta+r^6\tan^4\theta)}\end{aligned}\ ] ] > from @xmath170 and @xmath171 , we obtain the parameters @xmath172\label{beta } & & \beta = { ( r^2 - 1)\sin\theta\over\sqrt{r^4\tan^2\theta + 1}}\,.\end{aligned}\ ] ] we have plotted these parameters as a function of the angle @xmath99 in fig .
[ fig : alphabeta ] , for @xmath173 .
these parameters have maximum and minimum values given by @xmath174}\,,\hspace{5mm}\\[2 mm ] & & \beta_{\rm max } = { r^2 - 1\over r^2 + 1}\,,\end{aligned}\ ] ] as can be seen in fig .
[ fig : alphabeta ] . in this subsection we will impose the general bounds found in section iii to the double inflation model @xcite .
note that in this model it is assumed that the heavy field decays into cold dark matter , and therefore we only have one isocurvature component , cdi .
the bounds from cdi will be used to constrain this particular model .
we will leave for the future a detailed analysis of other two - field models of inflation . in order to derive specific constraints , it will be useful to take into account the following relation between @xmath2 and @xmath20 , see eqs .
( [ alpha ] ) and ( [ beta ] ) @xmath175 which corresponds to a straight line in the contour plot of fig .
3 . this way
, one can evaluate the likelihood at which a given value of @xmath168 is ruled out .
unfortunately , from the contours in fig . 3 , one can not restrict much the range of @xmath168 , except to exclude @xmath176 at 2@xmath121 . even if the model passes this constraint for a given @xmath168 , it is possible that the prediction on @xmath96 ( [ nad ] ) and @xmath116 ( [ niso ] ) do not agree with the bounds on these parameters . in our case ,
the bounds are so loose that any tilt is allowed . perhaps in the future , with better observational constraints , we may use the information on the tilts to further rule out double inflation models .
another model worth exploring is the particular one in which the two fields have equal masses , corresponding to a massive complex field @xmath177 .
if we rewrite @xmath178 , with modulus @xmath121 and phase @xmath16 , the lagrangian can be written as @xmath179 note that there is no potential for the phase , so it will be free to fluctuate , which will induce a large isocurvature component , as we will see , and which can be used to rule this model out . in this case , the curvature and entropy perturbations are @xmath180 & & { { \cal s}_{\rm rad}}(k ) = -{\kappa h_k\over\sqrt{2k^3}}{1\over\sqrt{s_k}}\ , e_s({\bf k})\,,\end{aligned}\ ] ] with @xmath181 and @xmath182 orthonormal .
therefore , @xmath183 the curvature perturbation has a tilt @xmath184 , but the isocurvature perturbation has no tilt , @xmath185 , and the two modes are uncorrelated , @xmath69 . for the moment , this model is not ruled out .
using the recent measurements of temperature and polarization anisotropies in the cmb by wmap , together with recent data from vsa , cbi , and acbar ; the matter power spectra from the 2df galaxy redshift survey and the sloan digital sky survey , as well as the recent supernovae data from the sn search team , one can obtain stringent bounds on the various possible isocurvature components in the primordial spectrum of density and velocity fluctuations .
we have considered correlated adiabatic and isocurvature modes , and find no significant improvement in the likelihood of a cosmological model by the inclusion of an isocurvature component , see table 1 and fig .
so , the pure adiabatic scenario remains the most economic and attractive scenario .
in contrast with the wmap analysis , we decided not to include any data from lyman-@xmath2 forests , since constraints on the linear power spectrum coming from these experiments are derived under the assumption of a plain adiabatic @xmath3cdm scenario .
we did not include either strong priors on the isocurvature spectral index , unlike in our previous work , and allowed this parameter to vary up to @xmath186 .
this conservative approach leads to a preference for models with a very blue isocurvature primordial spectrum , and to upper bounds on the isocurvature fraction significantly larger than in other recent analyses : on a pivot scale @xmath0mpc@xmath1 the amplitude of the correlated isocurvature component can be as large as about 60% for the cold dark matter mode , 40% for the neutrino density mode , and 30% for the neutrino velocity mode , at 2@xmath121 .
this leaves quite a lot of freedom , for instance , for double inflation models with two uncoupled massive fields . assuming that one of these fields decay into cold dark matter , our results simply imply that the mass of the heavy field can not exceed five times that of the light field at the 2@xmath121 confidence level .
it is expected @xcite that in the near future , with better data from planck @xcite and other cmb experiments , we will be able to reduce further a possible isocurvature fraction , or perhaps even discover it .
the present results also suggest that constraining the linear matter power spectrum on scales which are mildly non - linear today will also be crucial in this respect .
after completing this work , we came upon the paper of parkinson et al .
@xcite , where a similar analysis was done for a class of hybrid inflation models .
we thank andrew liddle for stimulating discussions .
m.b . thanks the group at sussex university for their warm hospitality and acknowledges support by a marie curie training site fellowship .
we also acknowledge the use of the cosmos cluster for our computations with the cosmomc code , and thank the sponsors of this uk - ccc facility , supported by hfce and pparc .
this research was conducted in cooperation with gsi / intel utilising the altix 3700 supercomputer .
this work was supported in part by a cicyt project fpa2003 - 0435 , and by a spanish - french collaborative grant bewteen cicyt and in2p3 .
99 a. d. linde , phys .
b * 158 * , 375 ( 1985 ) ; l. a. kofman and a. d. linde , nucl .
b * 282 * , 555 ( 1987 ) ; s. mollerach , phys .
b * 242 * , 158 ( 1990 ) ; a. d. linde and v. mukhanov , phys .
d * 56 * , 535 ( 1997 ) ; m. kawasaki , n. sugiyama and t. yanagida , phys .
d * 54 * , 2442 ( 1996 ) ; p. j. e. peebles , astrophys .
j. * 510 * , 523 ( 1999 ) .
d. polarski and a. a. starobinsky , phys . rev .
d * 50 * , 6123 ( 1994 ) ; m. sasaki and e. d. stewart , prog .
phys . * 95 * , 71 ( 1996 ) ; m. sasaki and t. tanaka , prog .
phys . * 99 * , 763 ( 1998 ) .
j. garca - bellido and d. wands , phys .
d * 53 * , 5437 ( 1996 ) ; * 52 * , 6739 ( 1995 ) . c. gordon , d. wands , b. a. bassett and r. maartens , phys
. rev .
d * 63 * , 023506 ( 2001 ) ; n. bartolo , s. matarrese and a. riotto , phys .
d * 64 * , 123504 ( 2001 ) .
d. wands , n. bartolo , s. matarrese and a. riotto , phys .
d * 66 * , 043520 ( 2002 ) .
f. finelli and r. h. brandenberger , phys .
d * 62 * , 083502 ( 2000 ) ; f. di marco , f. finelli and r. brandenberger , phys .
d * 67 * , 063512 ( 2003 ) .
d. langlois , phys .
d * 59 * , 123512 ( 1999 ) ; d. langlois and a. riazuelo , phys .
d * 62 * , 043504 ( 2000 ) .
g. efstathiou and j. r. bond , mon . not .
r. astron .
a * 218 * , 103 ( 1986 ) ; * 227 * , 33 ( 1987 ) ; p. j. e. peebles , nature * 327 * , 210 ( 1987 ) .
h. kodama and m. sasaki , int .
j. mod .
phys . a * 1 * , 265 ( 1986 ) ; * 2 * , 491 ( 1987 ) ; s. mollerach , phys . rev .
d * 42 * , 313 ( 1990 ) .
m. bucher , k. moodley and n. turok , phys .
d * 62 * , 083508 ( 2000 ) ; phys .
* 87 * , 191301 ( 2001 ) .
d. h. lyth and d. wands , phys .
b * 524 * , 5 ( 2002 ) ; d. h. lyth , c. ungarelli and d. wands , phys .
d * 67 * , 023503 ( 2003 ) .
t. moroi and t. takahashi , phys .
b * 522 * , 215 ( 2001 ) [ erratum - ibid .
b * 539 * , 303 ( 2002 ) ] ; phys .
d * 66 * , 063501 ( 2002 ) .
p. crotty , j. garca - bellido , j. lesgourgues and a. riazuelo , phys .
lett . * 91 * , 171301 ( 2003 ) . c. l. bennett _ et al .
_ [ wmap collaboration ] , astrophys . j. suppl . * 148 * , 1 ( 2003 ) ; d. n. spergel _ et al .
_ , astrophys .
j. suppl .
* 148 * , 175 ( 2003 ) ; h. v. peiris _ et al . _ ,
astrophys .
j. suppl .
* 148 * , 213 ( 2003 ) . c. l. kuo _ et al . _
[ acbar collaboration ] , astrophys .
j. * 600 * , 32 ( 2004 ) ; j. h. goldstein _ et al .
_ , astrophys .
j. * 599 * , 773 ( 2003 ) . j. a. peacock _ et al . _
[ 2dfgrs collaboration ] , nature * 410 * , 169 ( 2001 ) ; w. j. percival _ et al .
_ , mon . not .
r. astron .
a * 327 * , 1297 ( 2001 ) ; * 337 * , 1068 ( 2002 ) .
m. tegmark _ et al .
_ [ sdss collaboration ] , astrophys .
j. * 606 * , 702 ( 2004 ) .
r. stompor , a. j. banday and k. m. gorski , astrophys .
j. * 463 * , 8 ( 1996 ) ; p.
j. e. peebles , astrophys .
j. * 510 * , 531 ( 1999 ) ; e. pierpaoli , j. garca - bellido and s. borgani , jhep * 9910 * , 015 ( 1999 ) ; m. kawasaki and f. takahashi , phys .
b * 516 * , 388 ( 2001 ) ; k. enqvist , h. kurki - suonio and j. vliviita , phys .
d * 62 * , 103003 ( 2000 ) ; * 65 * , 043002 ( 2002 ) . c. gordon and a. lewis , phys
. rev .
d * 67 * , 123513 ( 2003 ) ; new astron .
rev . * 47 * , 793 ( 2003 ) .
j. valiviita and v. muhonen , phys . rev .
lett . * 91 * , 131302 ( 2003 ) .
k. moodley et al . , arxiv : astro - ph/0407304 .
see for instance , p. crotty , j. lesgourgues and s. pastor , phys .
d * 69 * , 123007 ( 2004 ) .
j. m. bardeen , phys .
d * 22 * ( 1980 ) 1882 ; v. f. mukhanov et al .
, phys . rep .
* 215 * ( 1992 ) 203 . | generic models for the origin of structure predict a spectrum of initial fluctuations with a mixture of adiabatic and isocurvature perturbations . using the observed anisotropies of the cosmic microwave backgound , the matter power spectra from large scale structure surveys and the luminosity distance vs redshift relation from supernovae of type ia
, we obtain strong bounds on the possible cold dark matter / baryon as well as neutrino isocurvature contributions to the primordial fluctations in the universe . neglecting the possible effects of spatial curvature and tensor perturbations
, we perform a bayesian likelihood analysis with thirteen free parameters , including independent spectral indexes for each of the modes and for their cross - correlation angle .
we find that around a pivot wavenumber of @xmath0mpc@xmath1 the amplitude of the correlated isocurvature component can not be larger than about 60% for the cold dark matter mode , 40% for the neutrino density mode , and 30% for the neutrino velocity mode , at 2 sigma . in the first case ,
our bound is larger than the wmap first - year result , presumably because we prefer not to include any data from lyman-@xmath2 forests , but then obtain large blue spectral indexes for the non - adiabatic contributions .
we also translate our bounds in terms of constraints on double inflation models with two uncoupled massive fields . |
in this paper we present the software tool .
first , we provide the motivation and a short description .
we then present the natural deduction system as it is done in a popular textbook @xcite and as is it done in by looking at its formalization in isabelle .
this illustrates the differences between the two approaches .
we also present the semantics of first - order logic as formalized in isabelle , which was used to prove the proof system of sound .
thereafter we explain how is used to construct a natural deduction proof .
lastly , we compare to other natural deduction assistants and consider how could be improved .
we have been teaching a bachelor logic course with logic programming for a decade using a textbook with emphasis on tableaux and resolution @xcite .
we have started to use the proof assistant isabelle @xcite and refutation proofs are less preferable here .
the proof system of natural deduction @xcite with the introduction and elimination rules as well as a discharge mechanism seems more suitable .
the natural deduction proof system is widely known , used and studied among logicians throughout the world .
however , our experience shows that many of our undergraduate computer science students struggle to understand the most difficult aspects .
this also goes for other proof systems .
the formal language of logic can be hard to teach our students because they do not have a strong theoretical mathematical background . instead
, most of the students have a good understanding of concrete computer code in a programming language .
the syntax used in isabelle is in many ways similar to a programming language , and therefore a clear and explicit formalization of first - order logic and a proof system may help the students in understanding important details .
formalizations of model theory and proof theory of first - order logic are rare , for example @xcite .
we present the natural deduction assistant with a formalization in the proof assistant isabelle of its proof system .
it can be used directly in a browser without any further installation and is available here : http://nadea.compute.dtu.dk/ is open source software developed in typescript / javascript and stored on github .
the formalization of its proof system in isabelle is available here : http://logic-tools.github.io/ once is loaded in the browser about 250 kb with the jquery core library no internet connection is required . therefore can also be stored locally .
we display the natural deduction proofs in two different formats .
we present the proof in an explicit code format that is equivalent to the isabelle syntax , but with a few syntactic differences to make it easier to understand for someone trying to learn isabelle . in this format , we present the proof in a style very similar to that of fitch s diagram proofs .
we avoid the seemingly popular gentzen s tree style to focus less on a visually pleasing graphical representation that is presumably much more challenging to implement .
we find that the following requirements constitute the key ideals for any natural deduction assistant .
it should be : * easy to use .
* clear and explicit in every detail of the proof .
* based on a formalization that can be proved at least sound , but preferably also complete .
based on this , we saw an opportunity to develop which offers help for new users , but also serves to present an approach that is relevant to the advanced users . in a paper considering the tools developed for teaching logic over the last decade @xcite ,
the following is said about assistants ( not proof assistants like isabelle but tools for learning / teaching logic ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ assistants are characterized by a higher degree of interactivity with the user .
they provide menus and dialogues to the user for interaction purposes .
this kind of tool gives the students the feeling that they are being helped in building the solution .
they provide error messages and hints in the guidance to the construction of the answer .
many of them usually offer construction of solution in natural deduction proofs . [ ... ] they are usually free licensed and of open access . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ we think that this characterization in many ways fits . while might not bring something new to the table in the form of delicate graphical features , we emphasize the fact that it has some rather unique features such as a formalization of its proof system in isabelle .
we consider natural deduction as presented in a popular textbook on logic in computer science @xcite .
first , we take a look substitution , which is central to the treatment of quantifiers in natural deduction
. the following definition for substitution is used in ( * ? ? ?
* top ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ given a variable @xmath0 , a term @xmath1 and a formula @xmath2 we define @xmath3 $ ] to be the formula obtained by replacing each free occurrence of variable @xmath0 in @xmath2 with @xmath1 .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the usual side conditions that come with rules using this substitution seem to be omitted . in (
* top ) , we are shortly after given the following definition of what it means that @xmath1 must be free for @xmath0 in @xmath2 : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ given a term @xmath1 , a variable @xmath0 and a formula @xmath2 , we say that @xmath1 is free for @xmath0 in @xmath2 if no free @xmath0 leaf in @xmath2 occurs in the scope of @xmath4 or @xmath5 for any variable @xmath6 occurring in @xmath1 .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the following quote ( * ? ? ?
* bottom ) from the same book emphasizes how it seems more preferable , due to the high level of complexity , to avoid the details of these important side conditions : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ it might be helpful to compare @xmath1 is free for @xmath0 in @xmath2 with a precondition of calling a procedure for substitution . if you are asked to compute @xmath3 $ ] in your exercises or exams , then that is what you should do ; but any reasonable implementation of substitution used in a theorem prover would have to check whether @xmath1 is free for @xmath0 in @xmath2 and , if not , rename some variables with fresh ones to avoid the undesirable capture of variables .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ we find that this way of presenting natural deduction proof systems leaves out some important notions that the students ought to learn . in our formalization such notions and their complications
become easier to explain because all side conditions of the rules are very explicitly stated .
we see it as one of the major advantages of presenting this formalization to students .
we now present the natural deduction rules as described in the literature , again using @xcite .
the first 9 are rules for classical propositional logic and the last 4 are for first - order logic .
intuitionistic logic can be obtained by omitting the rule _ pbc _
( proof by contradiction , called `` boole '' later ) and adding the @xmath7-elimination rule ( also known as the rule of explosion ) @xcite .
the rules are as follows : side conditions to rules for quantifiers : @xmath13 : @xmath14 can not occur outside its box ( and therefore not in @xmath15 ) .
@xmath16 : @xmath1 must be free for @xmath0 in @xmath2 .
@xmath17 : @xmath1 must be free for @xmath0 in @xmath2 .
@xmath18 : @xmath14 is a new variable which does not occur outside its box .
in addition there is a special copy rule @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ a final rule is required in order to allow us to conclude a box with a formula which has already appeared earlier in the proof . [ ... ] the rule ` copy ' allows us to repeat something that we know already . we need to do this in this example , because the rule @xmath19 requires that we end the inner box with @xmath20 .
the copy rule entitles us to copy formulas that appeared before , unless they depend on temporary assumptions whose box has already been closed . though a little inelegant
, this additional rule is a small price to pay for the freedom of being able to use premises , or any other ` visible ' formulas , more than once . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the copy rule is not needed in our formalization due to the way it manages assumptions .
as it can be seen , there are no rules for truth , negation or biimplication , but the following equivalences can be used : @xmath21 the symbols @xmath22 and @xmath23 are arbitrary formulas .
one of the unique features of is that it comes with a formalization in isabelle of its proof system .
the terms and formulas of the first - order logic language are defined as the data types and ( later abbreviated and , respectively ) .
the type represents predicate and function symbols ( later abbreviated ) .
@ l + + + truth , negation and biimplication are abbreviations . in the syntax of our formalization
, we refer to variables by use of the de bruijn indices . that is , instead of identifying a variable by use of a name , usually @xmath0 , @xmath6 , @xmath24 etc .
, each variable has an index that determines its scope .
the use of de bruijn indices instead of named variables allows for a simple definition of substitution .
furthermore , it also serves the purpose of teaching the students about de bruijn indices .
note that we are not advocating that de bruijn indices replace the standard treatment of variables in general .
it arguably makes complex formulas harder to read , but the pedagogical advance is that the notion of scope is exercised .
provability in is defined inductively as follows : @xmath27 { { \textsf{ok r a}}}{{\textsf{ok ( dis p q ) a } } ~~&~~ { \textsf{ok r ( p \ # a ) } } ~~&~~ { \textsf{ok r ( q \ # a)}}}\ ] ] @xmath30 { { \textsf{ok q a}}}{{\textsf{ok ( exi p ) a } } & ~ { \textsf{ok q ( ( sub 0 ( fun c [ ] ) p ) \ # a ) } } & { \textsf{news c ( p\#q\#a)}}}\ ] ] @xmath31 { { \textsf{ok ( exi p ) a}}}{{\textsf{ok ( sub 0 t p ) a}}}\ ] ] means that the formula follows from the list of assumptions and means that is a member of . the operator is between the head and the tail of a list .
checks if the identifier does not occur in the any of the formulas in the list and returns the formula where the term has been substituted for the variable with the de bruijn index . instead of writing we could also use the syntax @xmath33 , even in isabelle , but we prefer a more programming - like approach . in the types we use @xmath34 for function spaces .
the definitions of , and are as follow : @ l + + + + + + + + + + + + + + + + + + + @ l + + + + + + + + + + + + + + + + + + + to give meaning to formulas and to prove sound we need a semantics of the first - order logic language .
this semantics is defined in the formalization in isabelle , and it is thus not part of the tool itself .
we present the semantics below .
is the environment , i.e. a mapping of variables to elements .
maps function symbols to the maps they represent .
these maps are from lists of elements of the universe to elements of the universe .
likewise , maps predicate symbols to the maps they represent . is a type variable that represents the universe . in
can be instantiated with any type .
for instance , it can be instantiated with the natural numbers , the real number or strings .
@ l + + + + + + + + + + + + semantics e f g ( p ) = ( x. semantics ( % n. if n = 0 then x else e ( n 1 ) ) f g p ) + semantics e f g ( p ) = ( x. semantics ( % n. if n = 0 then x else e ( n 1 ) ) f g p ) most of the cases of should be self - explanatory , but the case is complicated .
the details are not important here , but in the case for it uses the universal quantifier ( ) of isabelle s higher - order logic to consider all values of the universe
. it also uses the lambda abstraction operator ( % ) to keep track of the indices of the variables .
likewise , the case for uses the existential quantifier ( ) of isabelle s higher - order logic .
we have proved soundness of the formalization in isabelle ( shown here as a derived rule ) : @xmath35{{\textsf{semantics e f g p}}}{{\textsf{ok p [ \,]}}}\ ] ] this result makes interesting to a broader audience since it gives confidence in the formulas proved using the tool .
we now describe the core features of from the perspective of the user .
that is , we uncover how to use to conduct and edit a proof as well as how proofs are presented . in order to start a proof
, you have to start by specifying the goal formula , that is , the formula you wish to prove . to do so
, you must enable editing mode by clicking the edit button in the top menu bar .
this will show the underlying proof code and you can build formulas by clicking the red symbol . alternatively , you can load a number of tests by clicking the load button . at all times , once you have fully specified the conclusion of any given rule , you can continue the proof by selecting the next rule to apply .
again you can do this by clicking the the red symbol . furthermore , allows for undoing and redoing editing steps with no limits .
all proofs are conducted in backward - chaining mode .
that is , you must start by specifying the formula that you wish to prove .
you then apply the rules inductively until you reach a proof if you can find one .
the proof is finished by automatic application of the rule once the conclusion of a rule is found in the list of assumptions . to start over on a new proof
, you can load the blank proof by using the load button , or you can refresh the page . please note that any unsaved work will then be gone .
in we present any given natural deduction proof ( or an attempt at one ) in two different types of syntax .
one syntax follows the rules as defined in section [ dedrules ] and is closely related to the formalization in isabelle , but with a redefined and more simple syntax in terms of learning .
the proof is not built as most often seen in the literature about natural deduction .
usually , for each rule the premises are placed above its conclusion separated by a line .
we instead follow the procedure of placing each premise of the rule on separate lines below its conclusion with an additional level of indentation .
@xmath36 { p \land ( p \rightarrow q ) \rightarrow q } { \infer { q } { \infer { p \rightarrow q } { \infer[^{(1 ) } ] { p \land ( p \rightarrow q ) } { } } & \infer { p } { \infer[^{(1 ) } ] { p \land ( p \rightarrow q ) } { } } } } \ ] ]
throughout the development of we have considered some of the natural deduction assistants currently available .
several of the tools available share some common flaws
. they can be hard to get started with , or depend on a specific platform . however , there are also many tools that each bring something useful and unique to the table .
one of the most prominent is panda , described in @xcite .
panda includes a lot of graphical features that make it fast for the experienced user to conduct proofs , and it helps the beginners to tread safely .
another characteristic of panda is the possibility to edit proofs partially before combining them into a whole .
it definitely serves well to reduce the confusion and complexity involved in conducting large proofs .
however , we still believe that the way of presenting the proof can be more explicit . in , every detail
is clearly stated as part of the proof code . in that sense
, the students should become more aware of the side conditions to rules and how they work .
another tool that deserves mention is proofweb @xcite which is open source software for teaching natural deduction .
it provides interaction between some proof assistants ( coq , isabelle , lego ) and a web interface .
the tool is highly advanced in its features and uses its own syntax .
also , it gives the user the possibility to display the proof in different formats .
however , the advanced features come at the cost of being very complex for undergraduate students and require that you learn a new syntax .
it serves as a great tool for anyone familiar with natural deduction that wants to conduct complex proofs that can be verified by the system .
it may , on the other hand , prove less useful for teaching natural deduction to beginners since there is no easy way to get started . in , you are free to apply any ( applicable ) rule to a given formula , and thus , beginners have the freedom to play around with the proof system in a safe way . furthermore , the formalized soundness result for the proof system of makes it relevant for a broader audience , since this gives confidence in that the formulas proved with the tool are actually valid .
in there is support for proofs in propositional logic as well as first - order logic .
we would also like to extend to more complex logic languages , the most natural step being higher - order logic .
this could be achieved using the cakeml approach @xcite .
other branches of logic would also be interesting , and the possibilities are numerous .
apart from just extending the natural deduction proof system to support other types of logic , another option is to implement other proof systems as well .
because the tool has a formalization in isabelle of its proof system , we would like to provide features that allow for a direct integration with isabelle .
for instance , we would like to allow for proofs to be exported to an isabelle format that could verify the correctness of the proofs .
a formal verification of the implementation would require much effort , but perhaps it could be reimplemented on top of isabelle ( although probably not in typescript / javascript ) .
we would like to extend with more features in order to help the user in conducting proofs and in understanding logic .
for example , the tool could be extended with step - by - step execution of the auxiliary primitive recursive functions used in the side conditions of the natural deduction rules .
so far only a small group of computer science students have tested , but it will be classroom tested with around 60 bachelor students in the next semester . currently the tool has no support for student assignments and automatic feedback and/or grading . the tool could be extended such that the students are evaluated and perhaps given a score based on the proofs they conduct .
it is not obvious how this could best be implemented .
we hope to find the resources for the development of such features but already now we think that the tool has the potential to be one of the main ways to teach logic in mathematics and computer science .
we would like to thank stefan berghofer for discussions about the formalization of natural deduction in isabelle .
we would also like to thank andreas halkjr from and andreas viktor hess for comments on the paper .
jrgen villadsen , anders schlichtkrull and andreas viktor hess .
meta - logical reasoning in higher - order logic . accepted at 29th international symposium logica , hejnice monastery , czech republic , 15 - 19 june 2015 .
olivier gasquet , franois schwarzentruber and martin strecker .
panda : a proof assistant in natural deduction for all .
a gentzen style proof assistant for undergraduate students .
lecture notes in computer science 6680 , 8592 .
springer 2011 . | we present a new software tool for teaching logic based on natural deduction .
its proof system is formalized in the proof assistant isabelle such that its definition is very precise .
soundness of the formalization has been proved in isabelle .
the tool is open source software developed in typescript / javascript and can thus be used directly in a browser without any further installation .
although developed for undergraduate computer science students who are used to study and program concrete computer code in a programming language we consider the approach relevant for a broader audience and for other proof systems as well . |
agile ( `` astrorivelatore gamma ad immagini leggero '' , tavani et al . 2000 and barbiellini et al .
2000 ) is the first mission of the program of small missions of the italian space agency ( asi ) .
the main goal of agile is to monitor the gamma - ray sky in the energy range between 30 mev and 50 gev , with a large field of view ( @xmath23 sr ) , good sensitivity , good angular resolution and good timing ( dead time lower than 100 @xmath0s for the gamma - ray detector ) .
agile is scheduled for launch by the beginning of 2003 in an equatorial @xmath2100 minutes orbit , for a nominal lifetime of 2 years .
it will use the asi base in malindi as a ground station
. the satellite mass will be about 200 kg ( @xmath265 kg of payload ) and the power available to the payload is about 65 w. the scientific telemetry will be able to transmit approximately 300 mbit of data to ground at each passage over the malindi ground station .
the agile payload ( mereghetti et al . 2000 ) is composed of a si tracker , containing 14 planes of si microstrip detectors ( 121 @xmath0 m pitch ) , interleaved with tungsten layers used as a photon pair converter ( each layer is 0.07 radiation length - @xmath3 ) . at the bottom of the si tracker a mini - calorimeter - two planes of csi bars , for a total on - axis radiation length of 1.5 @xmath3 - is in charge of the total absorption of the created pairs . the same detector ,
however , can be used for independent detection and triggering of high energy ( 300 kev - 200 mev ) transient events .
the hard x - ray section , superagile , is located on top of the si - tracker .
all the above parts are surrounded by an anticoincidence made of a 6-mm thick ( 5-mm on the top shield ) , segmented plastic scintillator .
superagile is basically composed of a detection plane ( dp ) , a collimator equipped with a coded mask , a front - end electronics and an interface electronics ( saie ) .
table 1 resumes the main superagile characteristics and figure 1 shows its appearance ( see also soffitta et al .
2000 for a more extensive description ) .
the dp is composed of 4 detection units ( dus ) , placed on a single al honeycomb plane support , so that two of them sample the x - direction and the other two are devoted to the y - direction .
each du is composed of 4 si microstrip tiles , bonded in pairs so that the effective length of each strip is approximately 19 cm . they are read - out through a set of ide as - xaa1 chips , based on asic technology , 12 for each of the dus .
the collimator ( present baseline : 500 @xmath0 m thick carbon fiber field separators , coated with a 75 @xmath0 m gold layer ) is mounted on the same tray supporting the dp , and in turn supports the 4 orthogonal , one - dimensional coded masks .
the coded masks have been designed based on an hadamard sequence , with a 50% covering factor .
they will be manufactured either in gold or in tungsten , 100 @xmath0 m thick .
the saie is in charge of interfacing superagile with the agile data handling system , allowing an event - by - event transmission with better than 5 @xmath0s timing resolution .
the energy information will be provided in the extended energy range between 1 and 64 kev , with 64 channels , to allow for a finer threshold calibration at low energies , and exploit for calibration purposes the tungsten fluorescences at @xmath258 kev .
the combined capabilities of the saie and the agile data handling ( see also morselli et al . 2000 ) will allow the transmission to ground of a relatively large set of scientific housekeeping data , including ratemeters and detector images .
in particular , the agile data handling will be able to perform a continuous automatic search for transient events ( e.g. , gamma - ray bursts ) on timescales from 1 ms to 100 s. once a transient event is triggered onboard , the data handling will be able to provide attitude - corrected sky images for it , determining the location of the transient source on the sky . the possibility to distribute in almost real time
the coordinates of a transient event through a fast link ( e.g. , tdrss or similar ) is currently under study . [ tab ] .characteristics of superagile [ cols="<,<",options="header " , ]
the most critical items for the superagile design were the signal - to - noise ratio at low energies ( i.e. , around 10 kev ) and the power consumption .
for this reason extensive laboratory tests have been performed on the asic chips planned to be used as front - end electronics : the xa1.3 chip , precursor of the xaa1 chip , especially developed by ide as for superagile .
the results of these tests ( see del monte et al . 2000 for a detailed discussion ) show that the electronic noise can be reduced to less than 4 kev ( fwhm ) and the power consumption to less than 1 mw per channel ( superagile includes 6144 independent electronic channels ) . the same tests have shown a critical dependence of several chip characteristics ( gain , offset , strip address and others ) from the temperature .
we studied the sensitivity and expected astrophysical performances by means of analytical calculations and monte carlo simulations . in figure 2
we present the map of the superagile effective area for two individual dus over their field of view ( fov ) , for a monochromatic 13.1 kev photon beam ( this energy corresponds to the peak in the area vs. energy relation ) .
the area of two orthogonal units is presented , showing the overlap of their fov , providing an effective bi - dimensional source location capability within the central 60@xmath4 x 60@xmath4 part of the fov , in addition to the one - dimensional location capability for the further @xmath220@xmath4 .
the maximum effective area at the center of the fov is @xmath280 @xmath1 , thus giving a maximum of 320 @xmath1 when the four units are considered together . in figure 3
the 5-@xmath5 sensitivity is presented for sources with crab - like energy spectra , for an integration time of 50 ks , over the full 10 - 40 kev energy range .
the sensitivity is presented as a function of the source location within the field of view , along the two directions .
it is worth noticing the wide central part of the fov with slowly variable sensitivity , allowing for an excellent use of most of the fov for source monitoring purposes .
in this section we present the expected performances of superagile for few interesting classes of hard x - ray sources . in figure 4
we show the sensitivity ( 5-@xmath5 in 50 ks ) as a function of energy . on the same plot
we show the typical energy spectrum of the x - ray binary pulsar her x-1 , as observed by bepposax near the maximum of the 35-day cycle ( dal fiume et al . , 1998 ) , showing that superagile will be able to provide accurate energy spectra for this and similar sources , also when it goes to its periodic minimum ( a factor @xmath23 fainter ) . as a result of the long pointings ( @xmath22 weeks ) driven by the gamma - ray tracker
, superagile will point the same source(s ) for the same long time , thus providing both accurate energy spectra over such a long integration time as well as a monitoring of the fluxes and energy spectra over much shorter timescales . in figure 4
we also show the observed energy spectra ( in a flare and in a dip state ) of the galactic microquasar grs 1915 + 105 ( feroci et al . 1999 ) .
also in this case the superagile sensitivity is perfectly suited to allow both spectral and temporal variability studies .
it is worth noticing that the galactic microquasars are among the primary targets also for the gamma - ray tracker and will likely be pointed several times over the agile lifetime . in figure 5
we show the superagile capability in studying fast hard x - ray transients , as the short recurrent bursts from the soft gamma - ray repeaters ( e.g. , aptekar et al .
the plot clearly shows that also in 250 ms superagile can provide detailed energy spectra of such events .
the wide field of view of superagile is very well suited for a _ monitoring _ of the activity of these sources , that are mostly concentrated towards the galactic center , thus allowing to trigger pointed observations when they undergo periods of intense bursting activity .
furthermore , the giant flares from these sources ( e.g. , feroci et al .
2000 ) are very good candidates for emission of rapid and intense flashes of gamma - rays and superagile can provide the tracker with an accurate position of a possible new soft gamma - ray repeater manifesting itself with a giant flare , as the source of the 1979 march 5@xmath6 event did . finally , figure 6 shows the source location capability of superagile for gamma - ray bursts .
the plot simulates the sky image of the superagile detection of grb 980425 , a relatively weak and soft gamma - ray burst in the bepposax sample .
although the simulation assumes an event at 15@xmath4 off - axis in both x and y directions , the detection is highly significant and allows for a very good source location determination , as can be seen by the right - hand panels where the sky image is zoomed - in .
this capability motivated the set - up of the onboard triggering and source localization system for superagile ( see sec .
aptekar , r. , et al .
2000 , astro - ph/0004402 barbiellini , g. , et al .
2000 , proc .
fifth compton symposium , aip 510 , p.750 dal fiume , d. , et al .
1998 , a&a 329 , l41 del monte , e. , et al .
2000 , proc .
spie conference 4140 feroci , m. , et al .
1999 , a&a 351 , 985 feroci , m. , et al .
2000 , apj in press ( astro - ph/0010494 ) mereghetti , s. , et al .
2000 , these proceedings morselli , a. , et al .
2000 , proc .
spie conference 4140 soffitta , p. , et al .
2000 , proc .
spie conference 4140 tavani , m. , et al .
2000 , proc .
fifth compton symposium , aip 510 , p.746 | superagile is the x - ray stage of the agile gamma - ray mission .
it is devoted to monitor x - ray ( 10 - 40 kev ) sources with a sensitivity better than 10 mcrabs in 50 ks and to detect x - ray transients in a field of view of 1.8 sr , well matched to that of the gamma - ray tracker , with few arc - minutes position resolution and better than 5 @xmath0s timing resolution .
superagile is designed to exploit one additional layer of four si microstrip detectors placed on top of the agile tracker , and a system of four mutually orthogonal one - dimensional coded masks to encode the x - ray sky .
the total geometric area is 1444 @xmath1 .
low noise electronics based on asic technology composes the front - end read out .
we present here the instrumental and astrophysical performances of superagile as derived by analytical calculations , monte carlo simulations and experimental tests on a prototype of the silicon microstrip detector and front - end electronics .
[ 1999/12/02 1.01 ( pwd ) ] |
layered iridium perovskites ( ruddlesden - popper series : sr@xmath0ir@xmath1o@xmath2 , where @xmath21 ) serve as an intriguing playground for exploiting the novel properties of @xmath22 electrons arising from competition between the spin - orbit ( so ) interaction , crystal field , and coulomb interaction ( electronic correlation ) of comparable energy in @xmath23 transition metal oxides .
recent studies with a variety of microscopic probes have established that the monolayer compound sr@xmath6iro@xmath7 is a mott insulator realized within @xmath24 manifolds that are reorganized by the so interaction : the @xmath24 band splits into two narrow subbands mainly composed of @xmath25 and @xmath26 spin - orbital states , where the latter is further split by the coulomb interaction ( @xmath27 ) into upper and lower hubbard bands upon half - filling.@xcite the bandwidth ( @xmath28 ) increases with the number of iro layers ( @xmath29 ) , which leads to the suppression of the mott transition ( @xmath30 ) and emergence of an anomalous metallic state in sriro@xmath8.@xcite regarding magnetic properties , sr@xmath6iro@xmath7 is known to exhibit weak ferromagnetism below the curie temperature @xmath31 k in accordance with the mott insulating phase , which is characterized by small saturation moments ( @xmath32/ir).@xcite this is now understood to come from a canted antiferromagnetic ( af ) order within the iro@xmath6 ( @xmath33 ) plane,@xcite where the spin canting is induced by the strong so interaction mapped onto the pseudo - dipolar interaction and the dzyaloshinsky - moriya interaction between the @xmath26 isospins.@xcite while similar behavior has been reported for sr@xmath8ir@xmath6o@xmath9 ( @xmath34/ir),@xcite recent studies using resonant x - ray scattering have shown that the magnetic structure is a g - type af@xcite with @xmath35-axis collinear moments,@xcite where the spin flop from sr@xmath6iro@xmath7 to sr@xmath8ir@xmath6o@xmath9 is attributed to the three - dimensional ( cubic ) character of the @xmath26 ground state that mediates the inter - layer pseudo - dipolar interaction common to the in - plane.@xcite however , these compounds exhibit complicated magnetic properties at low temperatures ( and under weak magnetic fields ) that do not necessarily fit into the current framework of effective theory .
recent studies on sr@xmath6iro@xmath7 have revealed a magnetic anomaly at low temperatures that is quite sensitive to an external field .
the magnetization is anomalously reduced below @xmath36 k at 0.05 t ( where @xmath37 is reduced to @xmath3850 k at 0.15 t ) , which suggests an enhanced antiferromagnetic correlation below @xmath37.@xcite the anomaly has also been shown to be closely linked with the charge sector ; the dielectric response exhibits crossover from relaxor - like to ferroelectric behavior as the temperature is decreased past @xmath37 .
these features correlate with the reduction in the ir - o - ir bond angle , which suggests a change in spin canting induced by alteration of the exchange coupling .
while the situation is not as clear for sr@xmath8ir@xmath6o@xmath9 , earlier reports in the literature have reported the observation of a rapid decease in magnetization at @xmath39 k and reversal in the basal plane below 20 k under field - cooled conditions at 0.01 t. this has only been observed at low magnetic induction ( below @xmath40 t).@xcite again , these anomalies exhibit a certain correlation with the charge sector as inferred from the sharp increase in the electrical resistivity.@xcite in contrast to these two cases , no clear anomaly suggesting a phase transition has been observed in the bulk properties of orthorhombic sriro@xmath8 ( @xmath16 ) over the temperature range from ambient to the liquid helium temperature . the electrical resistivity is metallic above @xmath3845 k , where it exhibits a minimum , and changes to localizing behavior at lower temperatures.@xcite in addition , a positive magnetoresistance has been observed below @xmath38170 k.@xcite through a band calculation considering the so interaction and optical conductivity , the carrier has been inferred to have a large effective mass due to electronic correlation.@xcite muon spin spectroscopy ( @xmath5sr ) is a unique probing technique for investigating the _ local _ magnetic structure and spin dynamics with extreme sensitivity , which is useful for studying magnetism arising from ir moments to which neutron diffraction techniques are difficult to apply because of the large neutron absorption cross - section of ir nuclei . in particular , once the muon site(s ) in the relevant compound is confirmed , the local magnetic field probed by the muon serves as a true measure of the order parameter because it is directly proportional to the magnitude of local magnetic moments .
note that the information gained from @xmath5sr is complementary to that from diffraction - based techniques ( x - ray and neutron scattering ) , as the latter corresponds to the information averaged over an entire crystal volume . in this paper
, we report @xmath5sr measurements on powder samples of sr@xmath0ir@xmath1o@xmath2 ( @xmath10 and @xmath4 ) under a zero external field that establish the development of a long - range magnetic order in sr@xmath6iro@xmath7 ( @xmath41 defined by onset " @xmath42 k ) and in sr@xmath8ir@xmath6o@xmath9 ( @xmath43 k ) .
more importantly , we demonstrate the presence of a secondary magnetic phase at lower temperatures ( below @xmath44 k in sr@xmath6iro@xmath7 and @xmath45 k in sr@xmath8ir@xmath6o@xmath9 ) that can be discerned by the increase in spontaneous local fields at muon sites located near the apical oxygen of the iro@xmath19 octahedron .
this result provides a microscopic basis for understanding the magnetic anomalies observed in the bulk properties of these compounds below @xmath18 and suggests that the order parameter of the secondary magnetic phase is not solely determined by the magnitude of ir local moments .
we show that the @xmath5sr result below @xmath18 is consistently reproduced by postulating the presence of small ordered moments at the apical oxygen in those two compounds , which is in line with the above anomalies in their bulk properties .
meanwhile , the absence of magnetism for @xmath16 supports the scenario that the @xmath24 band width is a monotonic function of the co - ordination number in this series of ir perovskites .
polycrystalline samples ( mosaic of small single crystals ) of sr@xmath6iro@xmath7 and sr@xmath8ir@xmath6o@xmath9 were prepared using a flux method ( with srcl@xmath6 as the flux ) , while the orthorhombic phase of sriro@xmath8 was obtained by synthesis under a high pressure of 5 gpa.@xcite structural analysis was performed using the powder x - ray diffraction technique to confirm that these samples mostly comprised a single phase with negligible content of impurities .
the present samples of sr@xmath6iro@xmath7 and sr@xmath8ir@xmath6o@xmath9 were prepared in the same batch as those measured by fujiyama _
et al_. using resonant x - ray diffraction.@xcite the bulk properties of these samples were investigated with different methods , including measurements of the magnetization , electrical resistivity , hall coefficient , and specific heat , to ensure the sample quality .
[ bchi ] shows the results of the magnetization measurements . ) of sr@xmath6iro@xmath7 and
sr@xmath8ir@xmath6o@xmath9 samples used for @xmath5sr measurements ( quoted from ref . ) .
( a ) @xmath46 vs. temperature , where measurements were made after zero - field cooling to 5 k with a sequence of 1 ) warming up to 350 k under a field @xmath47 t parallel to the @xmath48-axis and 2 ) then cooling down to 5 k. ( b ) @xmath46 vs. external field measured at 5 k ( @xmath49 ) . ] fig .
[ bchi]a clearly shows that a weak ferromagnetic behavior developed below @xmath38250 k in sr@xmath6iro@xmath7 , which is consistent with earlier reports.@xcite this is now understood to come from the canted ir moments that exhibit a metamagnetic transition at @xmath380.2 t.@xcite meanwhile , sr@xmath8ir@xmath6o@xmath9 exhibited much lower magnetization , although the temperature dependence was similar to that in sr@xmath6iro@xmath7 .
sr@xmath8ir@xmath6o@xmath9 crystals are known to occasionally contain iro@xmath6 monolayers ( corresponding to sr@xmath6iro@xmath7 ) as an impurity phase .
thus , the low magnetization in sr@xmath8ir@xmath6o@xmath9 may be attributed to such an impurity phase ; its fractional yield was empirically estimated to be less than 0.1% in single crystals.@xcite in contrast , the magnetic susceptibility in the orthorhombic sriro@xmath8 was found to be mostly independent of temperature ( @xmath50 emu / mol at 1 t , except for divergent behavior at low temperatures that may be attributed to paramagnetism associated with weak localization below @xmath3845 k ) .
the behavior is understood to be pauli paramagnetism.@xcite because the electronic specific coefficient is close to that of conventional metals ( @xmath51 25 mj / molk@xmath52 ) , the compound has a wilson ratio @xmath53 that is much greater than the limit of strong correlation ( @xmath54).@xcite such behavior is consistent with earlier reports@xcite and was attributed to a large van vleck term or strong magnetic correlation .
conventional @xmath5sr measurements were performed using the lampf spectrometer installed on the m15/m20 beamlines at triumf , canada . during measurements under a zero external field ( zf ) , the residual magnetic field at the sample position
was reduced to below @xmath55 t , while the initial muon spin direction was parallel to the muon beam direction [ @xmath56 . for longitudinal field ( lf )
measurements , a magnetic field was applied parallel to @xmath57 .
time - dependent muon polarization [ @xmath58 was monitored by measuring the decay - positron asymmetry along the @xmath59-axis : @xmath60 where @xmath61 is the average asymmetry . @xmath62\ ] ] is the positron event rate for the detector placed in the forward ( @xmath63 ) or backward ( @xmath64 ) position relative to the sample , @xmath65 is the muon decay lifetime ( @xmath66 s ) , @xmath67 is the decay positron asymmetry for the detector in question ( @xmath68 ) , and @xmath69 is the instrumental asymmetry [ @xmath70 under normal conditions ] .
[ tspec ] shows several examples of time - dependent zf / lf-@xmath5sr asymmetry spectra and their fast fourier transform ( fft ) in sr@xmath0ir@xmath1o@xmath2 ; those for @xmath3 and 2 exhibit a clear sinusoidal oscillation below @xmath71 .
such a signal indicates the onset of a relatively homogeneous internal field at muon sites in accordance with the long - range magnetic order .
the absence of enhanced muon depolarization just above @xmath71 ( usually expected as a consequence of the critical slowing down of the spin fluctuation ) and the volumetric expansion of fraction for the magnetically ordered region below @xmath71 ( as inferred from the increase in oscillation amplitude with decreasing temperature ) suggest that the transition is driven not by the magnetic interaction but by the mott transition .
in contrast , the spectra in sriro@xmath8 ( @xmath16 ) did not exhibit spontaneous oscillation over the entire observed range of the temperature , which demonstrates that no long - range magnetic order is present above 2 k. sr time spectra under zero external field ( zf ) in ( a ) sr@xmath6iro@xmath7 , ( b ) sr@xmath8ir@xmath6o@xmath9 and ( c ) sriro@xmath8 , with inset showing some spectra under longitudinal field ( lf ) .
modulation of sinusoidal pattern due to overlap of signals with different frequencies [ @xmath384.5 mhz and @xmath388 mhz at 1.9 k in ( a ) , @xmath389 mhz and @xmath3813 mhz at 1.7 k in ( b ) ] can be clearly observed in the earlier time range of the spectra up to 1 @xmath5s .
fast fourier transform of zf-@xmath5sr time spectra is shown for ( d ) sr@xmath6iro@xmath7 and ( e ) sr@xmath8ir@xmath6o@xmath9 .
( f ) lf-@xmath5sr spectra in sr@xmath8ir@xmath6o@xmath9 . ]
more interestingly , closer investigation of the spectra for the former two cases revealed a secondary frequency component at lower temperatures , which can be clearly observed in the fourier transform shown in fig .
[ tspec](d ) and ( e ) .
the signal amplitude of the secondary component increased instead of the first one , which suggests the gradual development of the former with decreasing temperature . considering this
, we analyzed the zf-@xmath5sr time spectra by curve - fitting using the following model function which is valid for the ordered phase of polycrystalline samples : @xmath72\nonumber\\ & & \hspace{9em}+a_{\rm para}g_z(t,\delta ) \label{fit}\end{aligned}\ ] ] where @xmath73 is the kubo - toyabe relaxation function@xcite to describe the slow gaussian depolarization due to random local fields exerted from _
nuclear _ magnetic moments ( with @xmath74 being the linewidth in the quasistatic limit ) , @xmath75 is the partial asymmetry for the magnetic phase ( with @xmath76 ) , @xmath77 is the asymmetry for the paramagnetic phase , @xmath78 is the longitudinal relaxation rate , @xmath79 is the transverse relaxation rate , @xmath80 is the muon spin precession frequency ( @xmath81 with @xmath82 mhz / t and @xmath83 being the local field at the muon site ) , and @xmath84 is the initial phase of precession .
note that the first term in eq .
( [ fit ] ) represents the situation where the kubo - toyabe depolarization would be quenched [ _ i.e. , _
@xmath85 for one third of the implanted muons exposed to a longitudinal internal field in the magnetic phase .
we found that four components ( @xmath86 ) were sufficient to describe the spectra observed in sr@xmath6iro@xmath7 .
similarly , the spectra in sr@xmath8ir@xmath6o@xmath9 required four components with one showing no oscillation , while one of the rest showed further splitting ( @xmath87 ) when @xmath88 .
fig . [ param ] shows the partial asymmetry ( @xmath75 , @xmath77 ) , muon precession frequency ( @xmath89 ) , and relaxation rate ( @xmath79 ) as determined from the curve - fitting of zf-@xmath5sr spectra in sr@xmath6iro@xmath7 as a function of temperature .
as discussed below , the splitting of signals into multiple frequencies can be attributed to the statistical distribution of implanted muons into magnetically inequivalent interstitial sites in the unit cell ( situated at the local minima of electrostatic potential ) , where muons are exposed to different internal fields .
the muon - electron hyperfine interaction is predominantly determined by the short - range magnetic dipolar interaction ; thus , it strongly depends on the local configuration of the nearest neighboring magnetic ions .
the onset temperature ( @xmath41 ) for the two precessing components @xmath90 ( @xmath91 for @xmath92 ) and @xmath93 agreed well with the @xmath94 k suggested by bulk magnetization.@xcite @xmath41 also perfectly matched with the values inferred from resonant x - ray diffraction ( rxd).@xcite curve - fitting of the frequency by the power law @xmath95 yielded the critical index @xmath96 ( see table [ tab214 ] for more detail ) which is situated somewhere between the three - dimensional ( 3d ) heisenberg ( @xmath97 ) and two - dimensional ( 2d ) ising ( @xmath98 ) spin systems . note that @xmath99 is distinctly smaller than that observed by rxd@xcite and neutron@xcite [ @xmath100 ( @xmath101 in ref . ) ] , which may be attributed to the volumetric expansion of magnetic domains with decreasing temperature as suggested by @xmath5sr [ see @xmath75 in fig .
[ param ] ] . the intensity of x - ray / neutron diffraction should be proportional to the product of @xmath89 ( @xmath102 the true order parameter ) and @xmath75 ( @xmath102 volumetric fraction of magnetic domains ) .
thus , precautions must be taken when interpreting diffraction data in terms of critical indices for the magnetic order parameter . , @xmath77 , ( b ) muon precession frequency @xmath89 , and ( c ) relaxation rate @xmath79 , @xmath103 in sr@xmath6iro@xmath7 .
solid curves are results of curve - fitting by @xmath104 , and dashed curves are visual guides . for the correspondence between @xmath89 and muon sites ,
see fig .
[ musite1 ] . ] nonetheless , the most salient feature in fig .
[ param ] is the gradual departure of @xmath105 from the curve predicted by the power law and further splitting into two frequencies ( @xmath106 ) below @xmath44 k ; it shows a much steeper increase as the temperature decreases below 2030 k. a corresponding behavior is suggested for @xmath93 , for which the relaxation rate increased below @xmath18 .
an additional component ( @xmath107 ) showing strongly damped precession ( @xmath108 ) also developed .
the signal was accompanied by a loss of the initial asymmetry ( seen from the reduction of @xmath109 by @xmath3810% , suggesting fast depolarization over a time range of @xmath110 s ) .
a similar @xmath5sr result was reported in a previous paper with some different details with regard to the behavior of fitting parameters.@xcite surprisingly similar behavior has previously been observed for the muon precession frequency in another ir perovskite ba@xmath6iro@xmath7 , except for the minor difference that the signal corresponding to @xmath93 is missing.@xcite the difference may be due to the difficulty in identifying the precession signal of relatively small amplitude ( @xmath111712% in sr@xmath6iro@xmath7 ) , although it may also be due to the change in the population of muon sites induced by the modulation of the crystal structure upon the substitution of sr by ba .
the zf-@xmath5sr spectra in sr@xmath8ir@xmath6o@xmath9 were described by four components ( @xmath1123 , 5 ) for @xmath113 , while another component ( @xmath107 ) appeared below @xmath18 .
such a two - step behavior is qualitatively parallel to that in sr@xmath6iro@xmath7 .
[ param2 ] shows the temperature dependence of @xmath75 , @xmath89 , and @xmath79 , where @xmath93 seems to be the primary component reflecting the magnetic order parameter for @xmath113 , while the behavior of the other components ( @xmath114 mhz , @xmath115 mhz , and @xmath116 mhz ) is relatively ambiguous . the transition temperature @xmath117 k determined by the onset of @xmath93 showed close agreement with the bulk properties , including the specific heat.@xcite curve - fitting of @xmath93 using @xmath118 yielded @xmath119 , which is closer to the value for the 2d - ising model ( @xmath120 ) than the case of sr@xmath6iro@xmath7 .
this suggests a stronger 2d characteristic of the magnetic order in sr@xmath8ir@xmath6o@xmath9 .
solid curve is fitting line , and dashed curves are visual guides . for the correspondence between @xmath89 and muon sites ,
see fig .
[ musite2 ] . ]
although the value of @xmath121 [ @xmath122 mhz ] happened to be close to @xmath121 in sr@xmath6iro@xmath7 [ @xmath123 mhz ] , the @xmath41 values determined by these signals clearly differed between the respective compounds .
( this in turn indicates that the @xmath93 component in the sr@xmath6iro@xmath7 sample can not be attributed to the sr@xmath8ir@xmath6o@xmath9 impurity phase . ) from the appearance of @xmath124 below @xmath125 k , sr@xmath8ir@xmath6o@xmath9 was inferred to undergo a secondary magnetic transition for @xmath126 .
the transition was accompanied by a reduction in @xmath127 , which suggests that the signal comes from muons formerly associated with @xmath90 for @xmath92 .
furthermore , both @xmath128 and @xmath129 exhibited a steep increase around @xmath18 , which is regarded as the critical divergence of the relaxation rate often observed for the conventional magnetic order . figs .
[ tspec](d ) and ( e ) show that the relative amplitude of @xmath124 ( @xmath130 ) was mostly independent of the temperature in sr@xmath8ir@xmath6o@xmath9 below @xmath18 .
this is in contrast to the case in sr@xmath6iro@xmath7 , where it increased as the temperature decreased below @xmath3830 k. these results suggest that these two phases ( or states " ) coexist in sr@xmath8ir@xmath6o@xmath9 below @xmath18 .
thus , while the magnetism of sr@xmath8ir@xmath6o@xmath9 shares the feature of a secondary phase below @xmath18 with sr@xmath6iro@xmath7 , the temperature dependence of its order parameter demonstrates one of the differences between these two compounds .
the above results are substantially different from that reported earlier in the literature.@xcite franke _ et al .
_ reported emergence of a single frequency component over a temperature range of @xmath131 k , where the magnitude of the frequency ( 1.82.8 mhz ) and its temperature dependence were rather close to that of @xmath90 in sr@xmath6iro@xmath7 .
in contrast , we found no corresponding signal in the present result for sr@xmath8ir@xmath6o@xmath9 .
they also refered to a signal component that exhibits extremely rapid relaxation ( @xmath3840 mhz ) over the entire temperature range that they studied ( @xmath132 k ) , which is missing from our result .
these observations suggest that the specimen used by franke _
et al._@xcite may have contained a significant fraction of ingrowth sr@xmath6iro@xmath7 phase upon preparation . unlike the preceding two cases , sriro@xmath8 remained non - magnetic over the entire temperature range studied ( 2 k@xmath133 k ) , which was inferred from the absence of a spontaneous muon precession signal in zf-@xmath5sr spectra .
more specifically , the spectra shown in fig .
[ tspec](c ) are the sum of two components showing gaussian depolarization due to nuclear random local fields . while the depolarization was mostly quenched by applying a longitudinal magnetic field of @xmath134 mt , a slowly depolarizing component remained [ see the inset of fig .
[ tspec](c ) showing the data at 2 k ] .
this suggests the presence of residual fast spin dynamics of ir 5@xmath22 electrons behind the random local fields from nuclear dipolar moments .
the spectra were analyzed by curve - fitting using the form @xmath135 which becomes @xmath136e^{-\lambda_it}\nonumber\ ] ] under a zero field , where @xmath137 is the gaussian relaxation rate determined by the second moment of nuclear dipolar fields and @xmath79 is the relaxation rate due to the fluctuation of paramagnetic moments .
we obtained @xmath138 and @xmath139 from the fitted curves , including those under lf-@xmath5sr measured at 2 k. fig .
[ tspec](c ) shows that the primary component of the time spectra ( @xmath140 ) did not appreciably change with time nor under different lf .
this makes it difficult to reliably determine @xmath141 and @xmath128 as independent parameters .
a similar difficulty arose for the secondary component from the small signal amplitude ( @xmath142 ) .
consequently , @xmath143 alone was treated as a parameter in the curve - fitting to determine the temperature dependence of the spectra , while @xmath144 was fixed to the values determined at 2 k ( except for @xmath145 k where @xmath146 was set free ) and @xmath128 was fixed to zero . in such a situation , @xmath143 extracted from curve - fitting would be equivalent to @xmath147 under the reverse conditions .
[ param3 ] summarizes the analysis results .
@xmath143 exhibited a weak enhancement around 3050 k , which may suggest a slowing down of paramagnetic spin fluctuation .
this is consistent with the bulk magnetic properties , which are mostly due to paramagnetism .
the electronic structure of sriro@xmath8 is reportedly on the verge of the metal - insulator transition , so the conducting carriers would have a large effective mass because of the strong electronic correlation.@xcite at this stage , there were a number of bulk property measurements with anomalies around 3050 k : the resistivity , seebeck , and hall coefficients exhibited a kink at @xmath3845 k , and the nernst effect showed a peak at @xmath3835 k. these anomalies seem to coincide with the enhancement of @xmath143 observed over a similar temperature range . and ( b ) relaxation rate @xmath137 , @xmath143 in sriro@xmath8 . ]
in the magnetically ordered states of sr@xmath6iro@xmath7 and sr@xmath8ir@xmath6o@xmath9 , the ir@xmath148 isospin state is virtually equivalent to the superposition of a magnetic dipole and octupole for @xmath149 electron systems , where the octupole makes a relatively small contribution to reducing the total magnetic induction .
thus , the local field at the muon site can reasonably be assumed to be predominantly determined by the point - like magnetic dipoles situated at ir sites .
the internal field @xmath11 at the muon site is given by the vector sum of the dipolar fields exerted from local ir@xmath148 isospin moments : @xmath150 where @xmath151 is the dipole tensor and is expressed by @xmath152 the summation runs through the @xmath153-th ir@xmath148 moment @xmath154 located at @xmath155 from a given muon site . for a given magnetic structure and ir@xmath148 moment size
, @xmath11 can be calculated for comparison with the observed @xmath156 the local electronic state associated with a muon is virtually identical with that of hydrogen as an interstitial atom because the difference in the reduced mass of the electron bound to the muon ( which determines the bohr atomic radius ) is only @xmath380.5% compared with that of hydrogen .
moreover , in transition metal ( tm ) oxides , the energy level of the electronic state associated with interstitial hydrogen @xmath157 has been empirically established to be predominantly determined by a common offset of nearly @xmath158 ev measured from the vacuum level regardless of the host compounds;@xcite the muon is predicted to stay diamagnetic ( accompanying no unpaired spin density ) as long as @xmath159 ( where @xmath160 is the energy at the bottom of the conduction band ) .
this is perfectly in line with the common observation that the muon in tm oxides satisfying the condition @xmath159 tends to form an o-@xmath5 covalent bond ( a muonic analogue of oh bond ) , where the local o-@xmath5 binding energy is the primary factor that determines @xmath157 . because sr@xmath0ir@xmath1o@xmath2 has a relatively narrow band gap ( i.e.
, @xmath160 is far below the vacuum level , so @xmath161 ) , we can safely presume that the muon forms an o-@xmath5 bond in these ir compounds .
accordingly , we searched for muon sites bound to oxygen ( with a typical o-@xmath5 distance of @xmath380.1 nm ) in the respective compounds that are situated near the ( local ) minima of the electrostatic potential that are induced by local ions .
the electrostatic potential was calculated using a custom - made computer program assuming a point charge for each atom,@xcite where the valence state and lattice structure were the input parameters .
in addition , the vienna ab - initio simulation package@xcite ( vasp , codes for the ab - initio quantum mechanical molecular dynamics calculation ) was used to examine the validity of the potential minima derived from the calculation based on the point - charge approximation .
these calculations showed perfect agreement for the case of sr@xmath6iro@xmath7 , while a slight difference was suggested for sr@xmath8ir@xmath6o@xmath9 and sriro@xmath8 ( see below ) . in the paramagnetic state ,
the muon - ir@xmath148 hyperfine parameter ( which is predominantly determined by the magnetic dipolar interaction ) is estimated by calculating the second moment @xmath162 ^ 2~~(\alpha = x , y,\:\beta = x , y , z ) , \label{dlt}\ ] ] where the sum runs over the @xmath69 and @xmath99 components of the terms exerting a magnetic field perpendicular to initial muon polarization from the @xmath153-th nuclear magnetic moment . the nuclear dipolar width ( @xmath163 )
can be estimated with the above equation by substituting @xmath154 with those of the nuclear magnetic moments . in the following
, the assignment of muon sites is discussed through a comparison of @xmath83 and the calculated values of @xmath11 in sr@xmath6iro@xmath7 and sr@xmath8ir@xmath6o@xmath9 .
in contrast , it is solely based on @xmath164 in sriro@xmath8 because the compound remained paramagnetic for the entire temperature range considered in the present study .
[ cols="^,^,^,^,^,^,^,^,^ " , ] the present model is also supported by the fact that , assuming an anitferromagnetic interaction between @xmath165 and @xmath166 ( as might be expected for the nearly @xmath167 o - ir - o bond angle ) , the reduction in the uniform magnetization below @xmath168 k in sr@xmath6iro@xmath7 and @xmath169 k and in sr@xmath8ir@xmath6o@xmath9 ( refs . ) can be qualitatively understood to be a result of its partial cancelation by oxygen polarization .
one note of interest is a recent example of the manganese compound tbmn@xmath6o@xmath170 , where an antiferromagnetic spin polarization at the oxygen sites was observed with oxygen @xmath171-edge resonant x - ray scattering.@xcite the compound exhibited an incommensurate antiferromagnetic order below 43 k that was associated with ferroelectric polarization because of the strong magneto - electric coupling .
it also showed the development of oxygen spin polarization below 38 k , where the dielectric constant exhibited a peak , and the origin of such multiferroic behavior was attributed to a charge redistribution around the ion nucleus in accordance with the alteration of the spin configuration , where spin polarization of the oxygen is a crucial component in the relevant theoretical model.@xcite the situation seems to be parallel with the case of sr@xmath6iro@xmath7 for which a similar anomaly in the dielectric constant was reported around 70 - 80 k , which is near @xmath18.@xcite
we used @xmath5sr to show that sr@xmath6iro@xmath7 and sr@xmath8ir@xmath6o@xmath9 exhibit clear signs of a two - stage magnetic order , whereas sriro@xmath8 stays a non - magnetic metal . provided that the predominant muon sites are located near the oxygen atoms common to these two compounds and that the effective ir moment size is 0.30.4@xmath172
, the magnitude of @xmath11 in the first stage order ( @xmath173 ) closely agrees with that predicted from the magnetic structures of each compound , which were recently inferred from resonant x - ray and neutron diffraction . for the secondary magnetic phase emerging below @xmath18
, we postulate that it originates from ordered magnetic moments ( @xmath380.03@xmath172 in sr@xmath6iro@xmath7 and @xmath380.09@xmath172 in sr@xmath8ir@xmath6o@xmath9 ) at the apical oxygen of the iro@xmath19 octahedron , which is induced by strong hybridization of ir @xmath22-o @xmath174 orbitals .
this hypothesis provides a qualitative explanation for the reported anomalies in both the magnetic and dielectric properties over the relevant temperature range according to the earlier literature .
a neutron diffraction study is currently underway to confirm the oxygen magnetic moments in those compounds .
the present results demonstrate the usefulness of the @xmath5sr technique in investigating local magnetism associated with oxygen , and it has the potential to be widely applicable to strongly correlated transition metal oxides owing to the fact that implanted muons tend to form a well - defined o-@xmath5 bonding state as interstitial pseudo - hydrogen atoms in oxides .
we express our gratitude to t. arima , s. fujiyama , y. murakami , h. nakao , h. okabe , and k. yamada for helpful discussion .
we also thank hua li for his help in calculating the electrostatic potential using vasp code and the triumf staff for their technical support during the @xmath5sr experiment .
this work was partially supported by the condensed matter research center , institute of materials structure science , kek .
99 b. j. kim , hosub jin , s. j. moon , j .- y .
park , c. s. leem , jaejun yu , t. w. noh , c. kim , s .-
oh , j .- h . park , v. durairaj , g. cao , and e. rotenberg , phys .
lett . * 101 * , 076402 ( 2008 ) .
i. franke , p. j. baker , s. j. blundell , t. lancaster , w. hayes , f. l. pratt , and g. cao , phys .
b * 83 * , 094416 ( 2011 ) .
h. okabe , m. isobe , e. takayama - muromachi , a. koda , s. takeshita , m. hiraishi , m. miyazaki , r. kadono , y. miyake , and j. akimitsu , phys .
b * 83*,155118 ( 2011 ) .
s. f. cox , j. l. gavartin , j. s. lord , s. p. cottrell , j. m. gil , h. v. alberto , j. piroto duarte , r. c. vilo , n. ayres de campos , d. j. keeble , e. a. davis , m. charlton , and d. p. van der werf , j. phys . : condens
. matter , * 18 * , 1079 ( 2006 ) . | in this study , the magnetic ground state of iridium perovskites ( sr@xmath0ir@xmath1o@xmath2 , where @xmath3 , 2 , and @xmath4 ) was considered using muon spin spectroscopy ( @xmath5sr ) . when probed by muons in sr@xmath6iro@xmath7 and sr@xmath8ir@xmath6o@xmath9 ( @xmath10 ) , the internal field ( @xmath11 ) showed clear sign of a magnetic order in two stages at transition temperatures @xmath12 k and @xmath13 k in sr@xmath6iro@xmath7 and @xmath14 k and @xmath15 k in sr@xmath8ir@xmath6o@xmath9 , respectively .
in contrast , no long - range magnetic order was observed in orthorhombic sriro@xmath8 ( @xmath16 ) .
based on the known magnetic structure in sr@xmath6iro@xmath7 and sr@xmath8ir@xmath6o@xmath9 , we successfully identified muon sites in these compounds from the magnitude of @xmath11 in the first stage ( @xmath17 ) .
below @xmath18 , @xmath11 probed by a fraction of muons occupying sites near the apical oxygen of iro@xmath19 octahedra exhibited a further increase but remained mostly unchanged for sites close to the in - plane oxygen . while such behavior can not be explained by the alteration of the ir spin structure , it is consistent with the selective appearance of ordered magnetic moments on the apical oxygen .
the oxygen polarization was also in line with the reported magnetization anomalies in these compounds below @xmath20 .
a possible link between the oxygen polarization and ferroelectric ( multiferroic ) behavior in sr@xmath6iro@xmath7 was considered . |
the scattering of medium - energy nucleons from nuclei can provide information about both nuclear structure and the @xmath4-interaction . in particular , the study of proton - nucleus scattering at intermediate energies is a useful method for determining accurate nuclear matter distributions in stable nuclei @xcite .
recently , it has also been used extensively to determine the extension of the nuclear matter density of exotic nuclei . in particular , in a recent set of experiments , inverse kinematics was used to study the elastic scattering of @xmath5he and determine their matter densities @xcite .
elastic scattering at intermediate energies can be calculated theoretically using either a non - relativistic @xcite or a relativistic optical @xcite model , usually with about the same good results .
an advantage of a relativistic optical model is that its two potentials simultaneously determine both the central and spin - orbit interactions .
when the relativistic impulse approximation ( ria ) is used @xcite , these interactions may be determined in terms of the corresponding scalar and vector nuclear densities @xcite . in this work
we use a relativistic nuclear optical potential constructed from a lorentz scalar , @xmath0 , and the time - like component of the four - vector potential , @xmath1 , to calculate elastic scattering angular distributions and analyzing powers for @xmath2he at intermediate energies and compare two different fits to the @xmath2he data and also to the ria results .
we also calculate angular distributions and analyzing powers of the elastic scattering of @xmath3he and compare an extension of the more physical of our two fits with the ria and the few available experimental data .
in relativistic optical model analyses of intermediate energy scattering @xcite , the dirac equation is usually used in the form , @xmath6+[v_{0}(r)+v_{c}(r)]\}\psi ( \vec{r})= e\psi ( \vec{r}),\ ] ] where @xmath0 is an attractive lorentz scalar potential , @xmath1 is the repulsive time - like component of a four - vector potential and @xmath7 is the coulomb potential .
the choice of the potentials is motivated by meson exchange considerations and simplicity . the simplest meson exchange interaction possessing a certain justification on physical grounds , which is also capable of providing nuclear saturation properties , takes into account the exchange of an attractive intermediate range isoscalar scalar meson and a repulsive shorter range isoscalar vector meson @xcite .
the two corresponding potentials @xmath0 and @xmath1 play an essential role in the description of both elastic scattering and polarization data at intermediate energies since their sum is the principal contribution to the central potential while their difference determines the spin - orbit interaction @xcite .
the scalar , @xmath0 , and vector , @xmath1 , optical potentials used in the analyses here can be written as @xmath8 with each of the potentials possessing both real and imaginary parts with possibly different radial dependences .
based on nuclear structure calculations , we take the radial dependence @xmath9 , for the case of @xmath10he , to have a gaussian form @xmath11.\label{eq3 } \end{aligned}\ ] ]
relativistic hartree calculations of the @xmath13he nucleus yield almost identical scalar and vector densities but an rms radius about 30% larger than the experimental one @xcite .
to obtain a physically reasonable ria potential for this system , we use equal scalar and vector matter densities having the rms radius of @xmath14 fm found in ref . @xcite together with the relativistic love - franey nn t - matrix of ref .
the results , represented by full lines in figs .
( [ fig1])-([fig2 ] ) , show that the ria agrees well with the experimental data at low momentum transfer but deviates substantially from the experimental results as the momentum transfer increases .
we also note that the ria angular distributions do not reproduce the oscilatory structure seen in the experimental data , decreasing monotonically instead .
the polarizations are not reproduced either , except at very low values of the momentum transfer . due to these differences ,
we have tried to fit the experimental data of ref .
@xcite using a dirac optical potential of the form given in equations ( [ eq2])-([eq3 ] ) . in a first attempt , labeled with the dashed lines in the figures ,
the parameters were left free to vary so as to obtain the simultaneous best fit to the experimental angular distributions and polarizations at the three values of the laboratory energy , @xmath15 . the same geometrical parameters were used at the three energies , while the strengths were assumed to vary linearly with the laboratory energy as @xmath16 , a common parametrization of optical model strengths .
the best fit parameters are given in table [ table1 ] .
as can be seen in the figures , we indeed obtain a fairly reasonable fit to the angular distribution and polarization at all three energies .
however , an interpretation of the imaginary part of the potential in the context of a ria potential would require a @xmath17 total cross section about 2.5 times the physical one , which makes the fit unsatisfactory on physical grounds .
the radii of the imaginary potentials are in agreement with the matter radius of ref .
@xcite , while the radii of the real potentials are found to be about @xmath18 bigger .
these also disagree with what one would expect from a ria potential , which usually yields real radii close to the matter radius and imaginary ones slightly larger . to obtain a more physically reasonable fit to the @xmath2he data
, we reduced the imaginary strengths of the first fit by a factor of 3 , reset the radii to the value of @xmath19 fm , corresponding to an rms radius of @xmath14 fm @xcite , and then let the parameters vary freely once again .
the results , labeled with dotted lines in the figures , present slightly poorer agreement with the experimental data , but continue to reproduce the oscillations in the angular distributions that are not obtained in the ria calculation ( full lines ) .
the imaginary potentials of the second fit are consistent with those of the ria potential , corresponding to a physically reasonable @xmath17 total cross section .
the radii are also closer to those of a ria potential .
.[table1]optical parameters for fits to the @xmath10he experimental data .
[ cols="^,>,>,>",options="header " , ] we next considered the elastic scattering of @xmath21he at laboratory energies of @xmath22 gev and @xmath23 gev , respectively . to perform the ria calculation for @xmath24he
, we followed the same procedure used for @xmath13he .
we generated equal scalar and vector matter densities using the appropriate gaussian harmonic oscilator ( ho ) geometrical parameters of ref .
@xcite , together with the relativistic love - franey nn t - matrix of ref .
the results of the ria calculation in figs .
( [ fig3 ] ) and ( [ fig4 ] ) agree well with the experimental data in the range over which the data exist .
we emphazise that the ria also describes the @xmath10he data well in this range of momentum transfers but , at larger momentum transfers , deviates substantially from the data in form and absolute values . for the cases of @xmath21he ,
the data is insufficient to attempt a fit .
we have thus attempted to extend our fits to the @xmath13he data by interpreting them in the context of a simple ria calculation .
to do so , we extract the effective strengths , @xmath25^{3/2 } $ ] , that , when multiplied by the @xmath13he density , would yield the potentials obtained in our fits to the @xmath10he data .
we have multiplied these strengths by the appropriate @xmath24he densities , obtained using the gaussian ho parametrization of ref .
@xcite to obtain effective dirac potentials for these systems , @xmath26 + \right .
\nonumber \\ & & \left.\zeta \frac{2}{3}\frac{1}{(\pi r_{0v}^{2})^{3/2}}\frac{r^{2}}{r_{0v}^{2 } } \exp\left[-r^{2}/r_{0v}^{2}\right]\right ) , \end{aligned}\ ] ] where @xmath27 for @xmath28he and @xmath29he , respectively .
the radii were modified slightly to take into account the differences between the experimental @xmath13he radius of ref . @xcite and those obtained in the fits .
our calculations for these two nuclei do not reproduce the experimental data as well as the ria calculation does , as can be seen in figs .
( [ fig3 ] ) and ( [ fig4 ] ) , which leads us to conclude that the parameters obtained in the @xmath10he fit can not be so easily extended to the @xmath3he elastic scattering
. however , we feel that they can transmit an idea of the discrepancies with an ria calculation that the experimental values could show . in figure ( [ fig4 ] )
we show the polarizations expected for proton scattering from these nuclei , for both the ria and our extension of the @xmath13he fits .
the differences between the two are again quite large . however , no data exist in this case . by comparision with the results for @xmath13he
it is difficult to say what one could expect of the experimental angular distribution and polarization in @xmath3he scattering at larger values of the momentum transfer .
we have calculated @xmath30he elastic scattering differential cross sections and polarizations using the relativistic impulse approximation and an adjusted dirac optical potential .
we have shown that the ria results obtained using the parameters of ref .
@xcite describe all three systems well at low values of the momentum transfer .
however , in the case of @xmath10he , for which more data exist , the ria deviates substantially from the data at higher momentum transfers . as an alternative to the ria we have ajusted the parameters of a dirac optical potential and then attempted to extend the parameters obtained to the cases of @xmath21he , but with unsatisfactory results . based on our analyses it is difficult to say
what one could expect at higher values of transferred momenta in the cases of @xmath21he .
more data are needed to elucidate this situation . in closing , we should mention that a possible reason for the failure of ria in describing the @xmath10he data is that it has been pushed beyond the limit of its applicability in the simple manner in which it is used here .
the decomposition of the effective dirac potential into a scalar potential and the fourth component of vector is valid in the target rest frame .
the scattering calculation , however , is performed in the cm frame and the boost that carries one frame to the other will convert the vector fourth component potential into a full vector potential . for proton scattering on a system such as @xmath31ca
the vector components introduced are small and can be neglected . for extremely light systems , such as those studied here ,
this is not the case .
we plan to examine the importance of this effect in the future .
g. d. alkhazov _ et al .
_ , _ phys . rep .
_ * c 42 * , ( 1978 ) , 89 .
g. d. alkhazov _ et al .
_ , _ phys .
lett _ * 78 * , ( 1997 ) , 2313 .
p. egelhof , _ progr .
phys . _ * 46 * , ( 2001 ) , 307 .
j. s. al - khalili and j. a. tostevin , _ phys .
_ * c 57 * , ( 1998 ) , 1846 .
k. amos , s. karataglidis and p. k. deb , _ nucl - th/0202081 _ , ( 2002 ) .
l. g. arnold , b. c. clark and r. l. mercer , _ phys .
_ * c 19 * , ( 1979 ) , 917 .
l. g. arnold , b. c. clark , r. l. mercer and p. schwandt , _ phys .
_ * c 23 * , ( 1981 ) , 1949 .
j. a. mcneil , j. r. shepard and s. j. wallace , _ phys .
lett . _ * 50 * , ( 1983 ) , 1439 , 1443 . b. d. serot and j. d. walecka , _ advances in nuclear physics _ , the relativistic many - problem , plenun press - new york ( 1986 ) .
k. erkelenz , _ phys .
* 13 c * ( 1974 ) , 194 .
g. brown and d. jackson , _ the nucleon - nucleon interaction _ , ( north - holland - amsterdam ) , 1976 .
t. ueda , f. e. riewe and a. e. s. grenn , _ phys .
_ * c 17 * , ( 1978 ) , 1763 .
w. h. furry , _ phys . rev . _
* 50 * , ( 1936 ) , 784 .
b. v. carlson and d. hirata , _ phys .
* c 62 * , ( 2000 ) , 054310 . c. j. horowitz , _ phys .
_ * c 31 * , ( 1985 ) , 1340 . | using a relativistic nuclear optical potential consisting of a lorentz scalar , @xmath0 , and the time - like component of a four - vector potential , @xmath1 , we calculate elastic scattering differential cross sections and polarizations for @xmath2he at intermediate energies for which experimental data are available .
we also calculate the differential cross sections and analyzing powers for @xmath3he at intermediate energies and compare with the few available experimental data . |
nuclear structure has been discussed mostly in the context of the liquid drop and shell models .
these models have been extremely useful in explaining many observed nuclear properties .
however , they are based on macroscopic concepts , and do not address the simplest nuclei , hydrogen and helium .
furthermore , recent @xmath22 experiments @xcite have indicated that in heavier near - closed - shell nuclei , less than @xmath23 of the nucleons are in the single - particle orbitals that would be fully occupied in the simple shell model . to obtain a more microscopic description of nuclear structure
, we may regard the nucleus as a collection of interacting nucleons described by the hamiltonian @xcite @xmath24 the interactions @xmath25 and @xmath26 are not exactly known , but @xmath25 is well constrained by the available scattering data , and binding energies and theoretical considerations place important constraints on @xmath26 .
the structure of the ground - state wave function @xmath27 at small interparticle distances is influenced by the repulsive core and tensor parts of @xmath25 .
most realistic models of @xmath25 contain these components and , for example , the reid @xcite , paris @xcite , urbana @xcite and new argonne @xmath28 @xcite models seem to predict similar structures .
the three - nucleon interaction @xmath26 is much weaker than the @xmath25 . due to a large cancellation between kinetic and two - body interaction energies
, it has a significant effect on nuclear binding energies @xcite but its effect on the structure of @xmath27 is much less than that of the better known @xmath25 . due to the strong spin - isospin dependence of @xmath25 and @xmath26 it is difficult to solve the schrdinger equation with the hamiltonian ( 1.1 ) .
only recently it has been possible to obtain accurate solutions for @xmath29 nuclei @xcite with the green s function monte carlo ( gfmc ) method .
accurate variational wave functions @xmath30 , which contain less than @xmath31 admixture of excited states are known for @xmath32 = 3 and 4 nuclei @xcite .
the available @xmath30 for @xmath6li @xcite and @xmath7o @xcite are certainly not as accurate as those for @xmath32 = 3 and 4 , nevertheless they presumably contain most of the important structure of the @xmath27 . in this paper
we examine the short - range structure of @xmath4h , @xmath5he , @xmath6li and @xmath7o by calculating the two - nucleon density distributions in states with isospin @xmath33 , spin @xmath1 , and spin projection @xmath2 . since the deuteron has only two nucleons , its one- and two - body density distributions are trivially related .
variational wave functions @xmath30 and monte carlo methods are used for @xmath34 .
the two - nucleon distributions in the @xmath20 states have a strong dependence on the spin projection @xmath2 .
the equidensity surfaces , spanning the top three quarters of the density range in @xmath2=0 states , have toroidal shape .
these tori are produced by the joint action of the repulsive core and tensor interactions .
in contrast the equidensity shapes in the @xmath2=@xmath31 have dumbbell shapes , which have been studied earlier in the deuteron @xcite .
a brief description of the two - nucleon interaction in @xmath20 states is given in section ii , and the density distribution of the deuteron is discussed in detail in section iii , where we show that the dumbbell - shaped distributions in @xmath2=@xmath31 states are produced by rotating tori .
commonly used models of @xmath25 predict that the maximum density torus has a diameter of @xmath171 fm , and the half - maximum density torus has a thickness of @xmath170.9 fm . in section iii
we relate these dimensions of the toroidal distribution to the observed electromagnetic form factors of the deuteron .
the structures are rather dense ; current models predict the maximum one - body density of the torus inside the deuteron to be @xmath35 @xmath36 , i.e. , approximately twice nuclear matter density .
the two - nucleon @xmath20 , @xmath37 density distributions in @xmath5he , @xmath6li and @xmath7o are compared with those of the deuteron in section iv .
the distributions for @xmath8 fm differ only by a single scale factor .
they indicate that in the @xmath0 state , the tensor correlations have near maximal strength for @xmath8 fm in all these nuclei .
the scale factor is identified as the levinger - bethe quasi - deuteron number , and its value is compared to the ratios of total photon ( @xmath38 = 80 to 120 mev ) and pion ( @xmath39115 mev ) absorption cross sections . in order to study the nature of many - body structures induced by these compact two - body structures we study
the @xmath9 , @xmath10 and @xmath11 overlaps with the @xmath30 of @xmath14he , @xmath15he and @xmath16li in section v. these depend strongly on the spin projection @xmath40 of the deuteron and indicate the presence of anisotropic structures in all these nuclei .
experiments to probe these structures are suggested .
pair distribution functions in other @xmath41 states are discussed in section vi .
those in @xmath42 , @xmath43 states are anisotropic as expected from the pion - exchange tensor force .
we also find that the number of @xmath44 pairs in a nucleus is reduced due to many - body effects involving the strong @xmath45 tensor correlations .
this reduction gives a significant contribution to the saturation of nuclear binding energies .
the skyrme field theory @xcite , related to qcd in the limit of large number of colors @xmath46 , has predicted toroidal shapes for the deuteron @xcite in the classical limit .
density distributions of the ground states with 3 to 6 baryons have also been calculated @xcite in this limit . in the last section , vii ,
we summarize our results , obtained with conventional nuclear many - body theory , and indicate their relation to those of the skyrme field theory in the classical limit , and of the constituent quark model .
nuclear forces are not yet quantitatively understood from qcd . however
, many realistic models have been constructed by fitting the available two - nucleon scattering data .
the shape of the short - range structures in the @xmath20 state appears to be relatively model - independent .
the interaction @xmath47 in the @xmath20 state in reid , urbana and argonne models can be expressed as @xmath48 while a @xmath49 operator is used instead of @xmath50 in the paris potential .
the structures are formed mostly by the static part of the interaction : @xmath51 it is instructive to study the expectation values of @xmath52 in eigenstates of the position operator @xmath53 with spin projections @xmath2=0 and @xmath2=@xmath31 .
these depend upon @xmath54 and @xmath55 , the polar angle of @xmath53 with respect to the spin - quantization axis @xmath56 , and are given by @xmath57 the @xmath2=0 expectation value of @xmath52 has the largest variation with @xmath55 as illustrated in fig .
[ potfig ] .
the static potential has a repulsive core ; outside the core it is very attractive for @xmath55=@xmath58 and repulsive for @xmath55=0 and @xmath59 .
therefore , in this state the @xmath60 pairs form a toroidal density distribution in the @xmath61 plane .
the potential in the @xmath2=@xmath31 states is attractive for @xmath55=0 and @xmath59 , and equal to that for @xmath2=0 , @xmath55=@xmath58 , while it is repulsive for @xmath55=@xmath58 and half way between the @xmath2=0 , @xmath55=0 and @xmath58 potentials .
thus the @xmath2=@xmath31 potential has two distinct minima separated by a barrier , and therefore the density distributions have a dumbbell shape in this state . at @xmath621.5 fm , the @xmath52
is dominated by the one - pion - exchange potential , while at smaller @xmath54 it has a significant model dependence .
much of this model dependence is cancelled by the differences in the momentum - dependent terms in the models .
in particular , the deuteron wave functions calculated from these potentials have much smaller model dependence .
these are commonly written as @xmath63 where @xmath64 and @xmath65 are the @xmath1- and @xmath13-state radial wave functions and @xmath66 are the spin - angle functions .
the @xmath67 and @xmath68 calculated from the different potential models are shown in fig .
[ r0r2fig ] .
the short - range structures are related to the @xmath67 and @xmath68 functions , and therefore we expect them to be fairly model - independent . however , the `` full '' bonn potential @xcite offers an exception .
the @xmath67 and @xmath68 predicted by this one - boson - exchange model of the @xmath69 interaction are similar to other predictions at larger @xmath54 , but they have an additional sharp structure with a range of @xmath170.2 fm .
we will not consider the possibility of such an additional structure in this work .
the short - range structure of the deuteron is most obvious in its density distribution @xmath70 which depends upon the projection @xmath40 of the total deuteron angular momentum , the distance @xmath71 from the deuteron center of mass , and the polar angle @xmath55 of @xmath72 ; it is independent of the azimuthal angle @xmath73 .
note that the interparticle distance @xmath74 , and the standard normalizations , @xmath75 & = & 1 , \\
\int\mbox{d}^3{\bf r}^{\prime}\ \rho_d^{m_d}({\bf r}^{\prime } ) & = & 2,\end{aligned}\ ] ] are used in this work .
the @xmath76 is given by @xmath77 , where the factor 16 comes from the difference in normalizations ( 3.1 ) and ( 3.2 ) .
a simple calculation using eq .
( 2.5 ) yields @xmath78 \>\>\ > , \\
\rho_d^{\pm 1}({\bf r}^{\prime})&=&\frac{4}{\pi}\left [ c_0(2r^{\prime})+c_2(2r^{\prime})\mbox{p}_2(\cos\theta ) \right ] \>\>\>,\end{aligned}\ ] ] with @xmath79 the interesting structure of these density distributions is shown in figs . [ figdeutxz]-[surffig ] .
[ figdeutxz ] shows @xmath80 along @xmath81 and @xmath82 directions , noting that @xmath83 the above densities are the largest while @xmath84 is the smallest as expected from the properties of @xmath52 discussed in sec .
the small value of the ratio @xmath85 indicates that the deuteron has near maximal tensor correlation at distance @xmath86 fm or equivalently at @xmath8 fm .
this ratio is @xmath170 for maximal tensor correlations .
figures [ rhod0 ] and [ rhod1 ] show the density distributions predicted by the argonne @xmath28 model in the @xmath87 plane .
the maximum value of @xmath88 is fairly model - independent ( fig .
[ figdeutxz ] ) and large ( @xmath170.35 @xmath36 ) .
the maxima of @xmath89 ( fig . [ rhod0 ] ) form a ring with a diameter of @xmath171 fm , denoted by @xmath90 , in the @xmath91 plane , while the @xmath92 has two equal maxima on the @xmath93-axis ( fig .
[ rhod1 ] ) at @xmath93=@xmath94 .
the three - dimensional distributions @xmath76 can be obtained by rotating the distributions shown in figs .
[ rhod0 ] and [ rhod1 ] about the @xmath93-axis .
they are represented by equidensity surfaces shown in fig .
[ surffig ] for @xmath88=0.24 and 0.08 @xmath36 ; all four sections are drawn to the same scale .
the maximum value of @xmath95 is @xmath170.05 @xmath36 ( fig .
[ figdeutxz ] ) .
therefore the equidensity surfaces for @xmath89 having @xmath960.05 @xmath36 can not intersect the @xmath93-axis , and thus have toroidal shapes shown in figs .
[ surffig]b and [ surffig]d . the central hole in these tori is due to the repulsive core in @xmath52 , and their angular confinement is due to the tensor force .
in absence of the tensor force , @xmath97 , the @xmath98 , and the equidensity surfaces are concentric spheres .
the maximum value of @xmath99 is @xmath170.19 @xmath36 as can be seen from fig .
[ figdeutxz ] .
therefore the equidensity surfaces of @xmath92 for @xmath1000.19 @xmath36 can not cross the @xmath55=@xmath58 plane ; they have two disconnected parts forming a dumbbell as shown in fig .
[ surffig]a . at smaller values of @xmath88
we also obtain two equidensity @xmath92 surfaces ( fig .
[ surffig]c ) , consisting of an inner surface due to the repulsive core enclosed by an outer . at
very small @xmath88 ( @xmath1010.05 @xmath36 ) the equidensity surfaces of @xmath89 also have disconnected inner and outer parts , neither close to spherical in shape .
the toroidal shape of the @xmath40=0 equidensity surfaces is more compact and persists down to smaller @xmath88 , or equivalently to larger values of @xmath71 , as can be seen from figs .
[ figdeutxz]-[surffig ] . in the classical skyrmion field theory only this shape
is obtained for the distribution of baryon density in the ground state for two baryons @xcite .
the deuteron can be considered to be more deformed in the @xmath40=0 state .
for example , the expectation values of the quadrupole operator @xmath102 obey the relation @xmath103 it is rather simple to obtain the @xmath104 from the @xmath105 .
we rotate the @xmath105 about the @xmath106-axis by an angle of @xmath58 so that the toroidal ring is in @xmath107 plane with @xmath108 as the symmetry axis .
this places the deuteron in the superposition of @xmath40=@xmath31 states .
the @xmath40=@xmath31 states are obtained by spinning the rotated toroid about the @xmath93-axis , and the @xmath104 is just the average value of the density of the spinning toroid , i.e. , @xmath109 the @xmath12=1 , @xmath110=@xmath31 states of the harmonic oscillator , given by @xmath111=e@xmath112i@xmath113 ) , are obtained in the same way from the @xmath111=e@xmath114 , @xmath12=1 , @xmath110=0 state .
therefore it is tempting to consider the toroidal shape of @xmath89 as the basic shape of the deuteron .
the expectation value of the current operator is zero in the @xmath40=0 state , therefore one may regard that as the `` static '' state of the deuteron .
note that the toroidal shapes can not be obtained by rotating the dumbbell by @xmath58 about the @xmath106-axis and spinning it about the @xmath93-axis .
this gives @xmath115 which is not true .
the dumbbell- or cigar - shaped density distribution of the deuteron in the @xmath40=@xmath31 state has been studied earlier @xcite .
unfortunately the toroidal distribution of the @xmath40=0 state was not studied , and its similarity with the predictions of skyrmion field theory was not noticed .
the deuteron electromagnetic structure functions @xmath116 and @xmath117 , and the tensor polarization @xmath118 in elastic electron - deuteron scattering have been extensively studied experimentally @xcite and theoretically @xcite .
they are usually calculated from the @xmath1- and @xmath13-wave functions @xmath67 and @xmath68 obtained from realistic interactions , by including in the nuclear electromagnetic current , in addition to the dominant impulse approximation ( ia ) operators , relativistic corrections and two - body meson - exchange contributions @xcite .
more recently , calculations of these observables based on quasipotential reductions of the bethe - salpeter equation and one - boson - exchange interaction models , constrained to fit nucleon - nucleon data , have also been carried out @xcite .
the theoretical predictions for the structure functions based on both the nonrelativistic and relativistic approaches are in good agreement with data .
our interest here is not to improve upon the present theoretical predictions , but to relate the values of the minima and maxima of @xmath118 and @xmath117 to the size of the toroidal structure in the deuteron .
the @xmath119-values of these extrema may be shifted by @xmath1710@xmath120 by corrections to the impulse approximation used in the following pedagogical analysis .
the charge form factors , defined as @xmath121 are shown in fig .
[ deutlong ] . at large @xmath119
the @xmath122 gets most of its contribution from the two peaks of @xmath123 ( fig .
[ rhod1 ] ) at @xmath93=@xmath94 .
the fourier transform of the sum of two @xmath124-functions at @xmath93=@xmath94 is given by @xmath125 with zeros at @xmath126=@xmath59 , 3@xmath59 , @xmath127 .
these zeros are due to the cancellation of the contribution from the two peaks , and they persist even when the peaks have a finite width .
the first two minima of @xmath128 , obtained from the argonne @xmath28 @xmath123 , occur at @xmath129=3.6 and @xmath130=12.6 @xmath131 .
the effective values @xmath132 estimated from the minima @xmath133 , using @xmath132=@xmath134 , are 0.87 and 0.75 fm for @xmath135=1 , 2 .
these values are smaller than the diameter @xmath90=1 fm because the dumbbell ends ( fig .
[ surffig]a ) are not spherical .
nevertheless the minima of @xmath136 seem to be primarily determined by the diameter @xmath90 of the maximum - density torus .
the fourier transforms of a disc of thickness @xmath137 , with @xmath138 perpendicular to the disc , are proportional to @xmath139 irrespective of the shape of the disc .
these have zeros at @xmath140=2@xmath59 , 4@xmath59 , @xmath127 , which may be used to obtain the thickness @xmath137 .
the first two zeros of @xmath141 at @xmath119=9.2 and 19.5 @xmath131 ( fig .
[ deutlong ] ) give values 0.68 and 0.64 fm for the effective thickness of the torus .
the maximum thickness along the @xmath142-axis of the calculated equidensity surface at half maximum density is 0.88 fm .
the @xmath118 form factor of the deuteron has small magnetic contributions which depend upon the electron scattering angle @xmath55 .
the extrema of @xmath143 are not significantly affected by this magnetic contribution as can be seen from fig .
[ deutt20 ] , and ignoring it we obtain a rather simple equation : @xmath144 its minima occur when @xmath128=0 , while the maxima have @xmath145=0 .
these minima and maxima correspond to those values of @xmath119 at which the recoiling deuteron has only @xmath40=0 or @xmath40=@xmath31 , respectively . the first minimum of @xmath143 is experimentally known to occur at @[email protected] @xmath131 in agreement with the value @xmath1471 fm predicted by realistic potentials .
the first maximum of @xmath143 has not yet been experimentally located ; it provides a measure of the thickness @xmath137 .
in magnetic elastic scattering the deuteron spin projection @xmath40 in the @xmath138 direction changes by @xmath31 since the photon has @xmath148=@xmath31 . thus the magnetic form factor
@xmath149 is a transition form factor .
it has convection current and spin - flip terms @xcite of which the latter is dominant .
the @xmath149 calculated with and without the convection current term are not too different ( fig .
[ deutmq ] ) .
the spin flip part of @xmath149 can be obtained from the transition density @xmath150 : @xmath151 \mbox{p}_2(\cos\theta)\right\}.\end{aligned}\ ] ] this transition density is shown in fig .
[ tranden ] ; it is dominated by the toroidal @xmath105 , and its effective thickness along @xmath93 axis , obtained from the zeros of @xmath149 ( fig .
[ deutmq ] ) , is @xmath170.85 fm .
the minimum of @xmath149 is observed @xcite at @xmath152 @xmath131 , supporting the theoretical estimates of @xmath137 .
the deuteron wave function in momentum space is defined as @xmath153 with @xmath154 the momentum distributions @xmath155 , given by @xmath156 , are then easily obtained as @xmath157 \>\>\ > , \\ { \tilde { \rho}}_d^{\pm 1}({\bf k})&=&\frac{1}{4 \pi}\left [ { \tilde { c}}_0(k)+{\tilde { c}}_2(k)\mbox{p}_2(\cos\theta_k ) \right ] \>\>\>,\end{aligned}\ ] ] where @xmath158 is the angle between @xmath159 and @xmath56-axis , and the @xmath160 are defined as in eqs . ( 3.5)-(3.6 ) with @xmath161 replaced by @xmath162 .
note that the @xmath163 are normalized such that @xmath164 the momentum distributions @xmath165 and @xmath166 for @xmath167 and @xmath168 are shown in fig .
[ rhodk ] .
note that @xmath169 the zeros of @xmath170 and @xmath171 occur at @xmath172 @xmath131 and @xmath173 @xmath131 , respectively , and are related to the spatial dimensions of the torus . in naive estimates
these minima occur at @xmath174 and @xmath175 , respectively . thus measuring the positions of the zeros in these momentum distributions
would provide an independent estimate of the spatial dimensions of the toroidal structure in the deuteron .
this information would be complementary to that yielded by elastic form factors measurements .
the momentum distributions @xmath176 could in principle be measured by @xmath22 scattering on polarized deuterons . in the one - photon exchange approximation the @xmath177
cross section , in the laboratory frame , is generally expressed as @xmath178 where @xmath40 is the target spin projection , @xmath179 and @xmath180 are the energy and solid angle of the final electron , and @xmath181 is the solid angle of the ejected proton .
the coefficients @xmath182 are defined in terms of the electron variables , while the structure functions @xmath183 involve matrix elements @xmath184 of the charge ( @xmath185 ) and current ( @xmath186 ) operators between the initial deuteron and final @xmath187 states .
the neutron momentum is @xmath188=@xmath189 , @xmath138 is the momentum transfer , @xmath190=@xmath188 is the missing momentum , and @xmath191 and @xmath192 are the proton and neutron spin projections , respectively @xcite . the cross section for unpolarized deuterons ,
@xmath193 has been measured up to @xmath194500 mev / c , and there is good agreement between theory and experiment @xcite . in plane - wave - impulse - approximation ( pwia ) , obtained by neglecting interaction effects in the final @xmath187 states as well as relativistic corrections and two - body terms in the charge and current operators ,
the @xmath40-dependent cross section is proportional to @xmath195 using , for example , tensor polarized deuterium , it should be possible to measure the difference between @xmath196 and @xmath197 , and therefore empirically determine the positions of the minima in these momentum distributions .
clearly , such an analysis is justified if the pwia is valid .
this assumption has been tested by carrying out the full and pwia calculations of the @xmath198 in parallel kinematics with @xmath119 fixed at 500 mev / c , @xmath199 in the range 290 - 390 mev , and the electron scattering angle @xmath200 .
the results , shown in fig
. [ 2heep ] , indicate that , while fsi , two - body current , and relativistic corrections are not entirely negligible , at least in the kinematical region which has been studied here , their effect is small compared to the difference between the cross sections for @xmath40=0 and @xmath31 .
we therefore conclude that the results of such an experiment could be used to empirically study the diameter and thickness of the torus .
one might argue that this information could be more easily obtained from elastic form factors measurements , as discussed above .
however , it should be realized that , in contrast to the @xmath201 data , the double - coincidence data would allow us to ascertain to what extent this toroidal structure is due to nucleonic degrees of freedom .
the two - nucleon density - distributions in @xmath33 , @xmath1 , @xmath202 , @xmath2 two - nucleon states are defined as @xmath203 here @xmath204 denotes the ground state of the nucleus with total angular momentum @xmath205 and projection @xmath206 , and @xmath207 projects out the specific two - nucleon state with @xmath208 . for @xmath209 nuclei ,
the wave functions used in this study are symmetric under exchange of neutrons and protons ; hence @xmath210 is independent of @xmath202 . for @xmath14he
, we have @xmath211 while for larger @xmath212 nuclei , the @xmath202 dependence is nontrivial . in the following we discuss @xmath213 , the average over @xmath202 of @xmath214 .
the @xmath213 is normalized such that @xmath215 which is the number of pairs in the nucleus .
it is a function of @xmath54 and @xmath55 independent of the azimuthal angle @xmath73 .
it can be verified from eqs.(3.3 - 3.6 ) and ( 4.1 ) that in the deuteron @xmath216 note that the spin - dependent two - body density on the left ( @xmath217 ) is an average over projections @xmath40 in the deuteron , while the polarized one - body density on the right ( @xmath218 ) has been summed over spins .
the @xmath213 in @xmath5he , @xmath6li , and @xmath7o have been calculated from variational wave functions using monte carlo techniques . for the a @xmath101 7 nuclei
, these wave functions minimize the expectation value of a hamiltonian consisting of the argonne @xmath28 two - nucleon and urbana ix three - nucleon potentials @xcite ( for @xmath32 = 6,7 the minimization is constrained by the experimental rms radii ) ; a detailed description of the form of the wave functions is given in refs .
the wave function for @xmath7o was obtained from the variationally best wave function by slightly increasing the radius of the single - particle part of the wave function so as to reproduce the experimental rms radius of @xmath7o .
the details of the @xmath7o wave function will be published elsewhere @xcite .
a cluster - expansion including up to four - body clusters with monte carlo integration @xcite was used to compute the two - body densities in @xmath7o . to reduce statistical fluctuations in the calculation , we write @xmath219 and
directly compute the @xmath220 as @xmath221 where @xmath222 represents the coordinates @xmath223 . because of the average over the total spin of the nucleus , the @xmath224 are zero for @xmath225 , and @xmath226 for the deuteron , the @xmath227 are related to the @xmath228 of eqs.(3.3 - 3.6 ) by @xmath229 ; @xmath230 .
the shapes of @xmath231 are very similar at @xmath232 2 fm in all the nuclei considered . in order to study the evolution of @xmath233 with @xmath32 we divided the @xmath234 by the ratio @xmath235 defined as @xmath236 the densities so normalized
are compared in fig .
[ rhoa ] , and the values of @xmath235 are listed in table i. fig .
[ rhoa ] shows @xmath237 , @xmath238 , and @xmath239 for @xmath4h , @xmath15he , and @xmath7o .
note that @xmath240 = @xmath241 by virtue of equations ( 4.4 ) , ( 4.6 ) and ( 4.7 ) in all nuclei .
after normalization by @xmath235 , the various densities for @xmath14he lie between those of @xmath4h and @xmath15he , while those for @xmath6li are in between the @xmath15he and @xmath7o results .
it is obvious from fig .
[ rhoa ] that the equidensity surfaces of the two - body density @xmath242 are very similar to those of the deuteron density shown in figures [ figdeutxz]-[surffig ] at @xmath243 2 fm ( @xmath244 1 fm ) . at @xmath243 2
fm the ratio @xmath245 is very small , indicating that the tensor correlations have near maximal strength in all the nuclei considered . in @xmath7o the @xmath233 becomes approximately independent of @xmath2 only for @xmath246 fm .
bethe and levinger suggested in 1950 @xcite that at small distances the relative @xmath18 = 0,1 neutron - proton wave function in a nucleus is likely to be similar to that in the deuteron .
we find that this is a good approximation .
the expectation value of any short - ranged two - body operator that is large only in the @xmath20 state scales as @xmath235 . in table
i we list values of the ratios of the calculated expectation values of the one - pion exchange part of the argonne @xmath28 potential , the observed low - energy ( 118 mev for @xmath14he @xcite and @xmath15he @xcite , and 115 mev for @xmath7o @xcite ) pion absorption cross sections and the average value of the observed photon absorption cross sections in the range @xmath38 = 80 to 120 mev .
all these processes are dominated by the @xmath20 pairs , and seem to scale as @xmath235 .
while comparing these ratios in detail it should be realized that @xmath247 in nuclei has a relatively small contribution from @xmath248 states , absent in the deuteron , which makes @xmath249 slightly larger than @xmath235 .
the @xmath250 = 21.3 mev for the argonne @xmath28 model , and it accounts for most of the deuteron potential energy , @xmath251 = 22.1 mev .
also in larger nuclei , the @xmath252 gives a large fraction of the total two - body interaction energy @xcite .
direct comparison of the ratio of pion absorption cross sections with @xmath235 may not be strictly valid .
the scattering and absorption of pions by spectator nucleons , absent in the deuteron , is expected to reduce the ratio @xmath253 , while three - body and higher absorption mechanisms , also absent in the deuteron , will increase it . after correcting for final state interactions of the two outgoing protons , the two - body ( @xmath254 ) part
is estimated to account for @xmath255 of the total absorption cross section for 115 mev @xmath256 by @xmath7o @xcite . in @xmath14he @xmath257 of the 118 mev @xmath256 absorption cross section
has three - body character @xcite , however a part of this @xmath258 must be due to initial and final state interactions . results of mainz experiments @xcite on @xmath259li and @xmath7o are used to calculate the average value of @xmath260 in the energy interval @xmath38 = 80 to 120 mev .
the @xmath261 averaged over the same energy interval is @xmath262 mb @xcite .
the only available data for @xmath14he in this energy range are from the experiments done in the 1960 s @xcite and 1970 s @xcite .
the average cross section of the two - body photodisintegration of @xmath14he , in the energy range 80 - 120 mev , is @xmath170.03 mb @xcite , and that for the three - body process is @xmath170.10 mb @xcite , giving total cross section of @xmath170.13 mb . the average cross section for total absorption of photons by @xmath15he in the same energy range is crudely estimated from fig .
[ potfig ] in ref .
@xcite to be @xmath170.3 mb .
the total number of pairs with given @xmath18 in nuclei can be computed as @xmath263 the values for @xmath20 and the corresponding naive independent - particle model values are also shown in table i. we see that the correlations induced by the potentials do not significantly change the @xmath264 from their independent - particle ( ip ) values ; however , as will be discussed later , this is not true for @xmath33=1 pairs . for few - body nuclei , @xmath235 is significantly larger than @xmath264 , however , in @xmath7o @xmath264 has a large contribution from pairs with large @xmath54 and @xmath235 is smaller than @xmath264 .
the calculated value of @xmath235 for @xmath7o is much smaller than levinger s estimate @xmath265 @xcite .
the @xmath266 has its half - maximum value at @xmath2671.8 fm ( fig .
[ rhoa ] ) .
if we identify the region with @xmath268 fm as the `` quasi - deuteron '' , then the probability that the @xmath60 pair in a deuteron is in the quasi - deuteron region is @xmath170.25 , and the number of quasi - deuterons in a nucleus is @xmath269 . in the past , however , @xmath235 itself has been interpreted as the number of quasi - deuterons in the nucleus .
the strong spin - dependent anisotropies of the two - nucleon densities suggest that three - nucleon and higher distribution functions in nuclei could also be anisotropic .
a general study of these higher distributions is beyond the scope of this work ; however the two - cluster distribution functions @xmath270 in @xmath14he , @xmath271 in @xmath15he , and @xmath272 in @xmath16li are simple to study with the monte carlo method @xcite .
they provide some information on the higher distribution functions , and may be relatively accessible by @xmath273 and @xmath274 experiments . the two - cluster overlap function can be written generally as @xmath275 where @xmath276 is the is the relative coordinate between the centers of mass of the two clusters and @xmath277 is an antisymmetrization operator for the two - cluster state .
the @xmath278 radial functions can be evaluated from @xmath279^{\dagger } y^*_{lm_l}(\hat{r}_{ab } ) \frac{\delta(r - r_{ab})}{r^2_{ab } } \psi^{m_j}({\bf r } ) \ , \end{aligned}\ ] ] where @xmath280 represents the coordinates of particles in cluster @xmath281 .
we note that in pwia the @xmath282 cross section is proportional to the momentum distribution @xmath283 obtained from the fourier transform of the overlap function @xmath284 . in the present work ,
the integrations have been made with monte carlo techniques akin to ref.@xcite , but with some improvements .
configurations are sampled with the weight function @xmath285 containing the full variational wave function . in ref.@xcite only a single term in the antisymmetric product in eq .
( 5.2 ) is calculated .
the efficiency of the monte carlo sampling has been improved by evaluating all possible partitions of the nucleus into clusters @xmath286 and @xmath287 at each configuration @xmath222
. we also use a much larger sample of configurations than in the previous calculations .
we can define a two - cluster wave function , in analogy with the deuteron wave function of eq .
( 2.5 ) , using the radial overlap functions @xmath288 where @xmath289 and @xmath290 denote spin states @xmath291 and @xmath292 of @xmath286 and @xmath287 . for the cases
@xmath293 = @xmath9 , @xmath10 , and @xmath11 there are both s- and d - wave states in the two - cluster wave function . in these cases the well known @xmath294 parameter can be defined by means of the @xmath295 and @xmath296 radial functions @xcite : @xmath297 although in the present paper we emphasize the short - range structure of nuclei , it is also interesting to study the asymptotic behavior of the overlap radial functions .
of particular interest is the asymptotic d / s ratio @xmath298 , where @xmath299 and @xmath300 are the asymptotic normalization constants of @xmath295 and @xmath296 respectively : @xmath301 here @xmath302 is the spherical hankel function of first kind and @xmath303 is the wave number associated with the separation energy of the nucleus into clusters @xmath286 and @xmath287 .
we must point out that the present variational method , as well as the gfmc method , determine the wave functions by energy minimization , to which long - range configurations contribute very little .
therefore these methods are not very sensitive to the asymptotic part of the wave functions , and consequently our values for @xmath304 should be considered only as estimates .
the two - cluster density distribution for a given set of spin projections is defined as : @xmath305 in each of the cases studied here , it exhibits spin - dependent spatial anisotropies which are easily understood in terms of the toroidal or dumbbell structure of the polarized deuteron .
the density is enhanced in the direction corresponding to the most efficient or compact placement of the deuteron with the remaining cluster , and reduced in those directions that would lead to very extended structures . finally , we are also interested in the total normalizations @xmath306 of the s- and d - wave two - cluster distributions : @xmath307 these quantities can be related to spectroscopic factors and give the total s- and d - state fractions .
all the results presented here are obtained from the argonne @xmath28 two - nucleon and urbana model ix three - nucleon interactions .
the calculated @xmath308 and @xmath309 are shown in fig .
[ r02dp ] ; the @xmath309 is negative and smaller in magnitude than the @xmath68 in deuteron ( fig .
[ r0r2fig ] ) .
the @xmath310 value obtained with these radial functions is @xmath311 fm@xmath4 , a little smaller than experimental estimates , ranging from @xmath313 to @xmath314 , obtained through dwba analysis of ( @xmath90,@xmath14he ) transfer reactions @xcite . in fig .
[ r02dp ] we also show our asymptotic fit to the s- and d - waves .
the result is @xmath315 , somewhat smaller than the faddeev result @xcite , @xmath316 .
experimental estimates , also obtained through dwba analysis of ( @xmath90,@xmath14he ) transfer reactions , range from @xmath317 to @xmath318 @xcite .
the total normalizations are @xmath319 and @xmath320 .
the sum , 1.33 , can be interpreted as the number of deuterons in @xmath14he @xcite .
it is less than 1.49 ( table i ) , the number of @xmath45 pairs , because the pairs are not always in the deuteron state .
it is also smaller than @xmath321 inferred from short range distribution functions ( fig .
[ rhoa ] and table i ) .
this is probably because @xmath14he is more compact than the deuteron .
the @xmath322 and @xmath323 are shown in figs .
[ 3hemd0 ] and [ 3hemd1 ] . when @xmath324 , the @xmath325 proton is preferentially along the @xmath142-axis ; in contrast , when @xmath326 , the @xmath327 proton is more likely to be in the @xmath61 plane . in the first density distribution the s- and d- wave amplitudes interfere constructively , to enhance the probability of finding the proton along the @xmath142-axis , whereas in the latter the interference is constructive on the transverse @xmath61-plane . consequently @xmath67 and @xmath68 have opposite signs and @xmath310 and @xmath328 are both negative .
the spin - dependent @xmath329 anisotropies are favored by both tensor and central forces , and lead to more compact three - body states .
the momentum distribution of @xmath9 clusters in @xmath330 @xmath14he is shown in fig .
[ 3hemdk ] for @xmath40 , @xmath191 = 0 , @xmath331 and 1 , @xmath332 for momenta parallel and transverse to the @xmath142-axis .
in pwia the @xmath333 cross - section is directly related to these momentum distributions .
a large spin dependence of the missing - momentum distribution for protons ejected in parallel kinematics is predicted for @xmath138 parallel to @xmath56 .
the minimum for the momentum distribution along @xmath142-axis for @xmath325 occurs at @xmath171.4 @xmath131 , while that for @xmath327 is at @xmath172.4 @xmath131 .
thus the spin asymmetry , @xmath334 , of the protons ejected from polarized @xmath14he changes from @xmath17 1 to + 1 as the missing momentum varies from @xmath171.4 to 2.4 @xmath131 in parallel kinematics and pwia .
the @xmath9 momentum distribution in unpolarized @xmath14he has been studied at saclay @xcite up to @xmath172.5 @xmath131 . the observed distribution is generally smaller than the pwia prediction @xcite indicating attenuation due to fsi .
however , a part of the fsi attenuation will cancel in the asymmetry , and moreover , it is now possible to perform continuum faddeev calculations @xcite including fsi .
the calculated @xmath335 and @xmath336 are shown in fig .
[ r02dd ] .
the @xmath337 value obtained with these radial functions is @xmath338 fm@xmath4 . in fig .
[ r02dd ] we also show our asymptotic fit to the s- and d - waves .
the result is @xmath339 .
the integrals of these functions yield @xmath340 and @xmath341 .
the number of deuterons present is greater than twice the sum of these quantities , 2.0 , when one allows for the additional presence of @xmath342 configurations .
the @xmath343 and @xmath344 are large in @xmath15he and their anisotropies , induced by the tensor interaction and the shapes of deuterons , are shown in fig .
[ 4hedd ] .
the @xmath343 is largest when @xmath345 is along the @xmath142-axis , i.e. , when the deuterons are in the toroidal shape and have a common axis .
it is smallest when @xmath346 is transverse ( two tori side by side ) and equal to that for @xmath344 with @xmath346 parallel to @xmath56 ( two dumbbells in a line ) .
the latter distribution is of intermediate strength when @xmath346 is transverse ( two dumbbells side by side ) . again in the first ( second ) density distribution the s- and d - wave amplitudes interfere constructively ( destructively ) along the @xmath56-axis .
therefore @xmath67 and @xmath68 have opposite signs and @xmath337 and @xmath347 are both negative .
the momentum distributions are also anisotropic ( fig .
[ 4heddk ] ) .
in particular the @xmath348 has a dip at @xmath3491.7 @xmath131 that is absent in the @xmath350
. it may be possible to study these with @xmath273 reactions .
the unpolarized @xmath15he@xmath351 reaction has been studied at nikhef @xcite .
the observed cross sections are much smaller than estimates using the @xmath10 momentum distribution and either pwia or dwia .
the calculated @xmath352 and @xmath353 are shown in fig .
[ r02ad ] . the @xmath352 and @xmath353 both exhibit nodes at short distances and have opposite signs almost everywhere .
this nodal structure has been predicted in @xmath354 and @xmath355 cluster models , but not always with the correct relative sign @xcite .
the asymptotic behavior is correlated with the quadrupole moment @xmath356 ; obtaining the experimental value of @xmath357 fm@xmath4 has been a long - standing problem in @xmath355 cluster models .
the variational wave function used here gives @xmath358 fm@xmath4 , i.e. , the correct sign but far too large in magnitude .
small changes in the long range part of the @xmath16li wave function have effects of order 1 fm@xmath4 on the quadrupole moment .
thus the values of the asymptotic properties , @xmath359=0.29 fm@xmath4 and @xmath360=0.07@xmath361 , obtained with this wave function may not be very accurate .
the total normalizations are @xmath362 and @xmath363 .
the resulting spectroscopic factor , 0.84 , is in good agreement with the value of 0.85 obtained in radiative capture experiments @xcite .
the two - cluster densities @xmath364 , multiplied by @xmath365 , are shown in fig .
[ 6liad ] .
they have two peaks ; the smaller inner peak at @xmath366 fm is almost spherically symmetric , while the larger peak at @xmath367 fm is anisotropic .
in particular @xmath368 is much larger than @xmath369 for @xmath370 fm . in the former configuration
the @xmath54 is along the axis of the torus , while in the latter it is transverse .
this anisotropy is also a consequence of the toroidal shape of the deuteron in the @xmath40=0 state .
in this section we discuss the properties in nuclei of pairs of nucleons with @xmath371 , @xmath372 , and @xmath373 . like the @xmath20 channel discussed in the previous sections , the @xmath19 channel also has a tensor potential , but it has the opposite sign of that for @xmath20
. therefore the role of @xmath2 is reversed compared to that in @xmath0 states ; @xmath374 pairs have maximum density along the @xmath142-axis , while @xmath375 pairs have maximum density in the @xmath61 plane as can be seen in fig .
[ ts11fig ] , which shows @xmath376 for @xmath15he , @xmath16li , and @xmath7o .
the curves for @xmath377 , @xmath378 are between the two sets of curves shown in the figure ; to reduce clutter they are not shown .
the curves for @xmath15he and @xmath16li have been renormalized by the factors @xmath379 to have the same peak height as for @xmath7o ; these factors are shown in table ii .
we see that the shapes of the @xmath19 density profiles are quite different in the different nuclei and that the anisotropy decreases as the number of nucleons increases .
the analog of the deuteron in the @xmath21 channel is the @xmath380s@xmath381 virtual bound state ( vbs ) . for the argonne @xmath28 potential , this is a pole on the second energy sheet at @xmath382 mev or @xmath383 @xmath131 .
although the wave function is not normalizable , it has a local peak which we scale to compare to the unpolarized deuteron in fig .
[ ts10fig ] .
we see that it peaks at a slightly larger radius and is broader .
the figure also shows the @xmath384 of @xmath15he , @xmath16li , and @xmath7o ; the curve for @xmath14he is between those of @xmath15he and @xmath16li , while the curve for @xmath259li is very close to that of @xmath16li .
again the curves have been normalized to the peak height of the @xmath7o density .
we see that the short - range shapes of the @xmath385 in nuclei are well reproduced by the vbs density .
finally , fig .
[ ts00fig ] shows the densities for the @xmath371 channel , again normalized to @xmath7o . as is the case for @xmath19 ,
there is no common shape .
table ii also shows the number of pairs , @xmath386 , in these @xmath18 channels and the corresponding ip values .
as is the case for @xmath20 ( table i ) , the number of pairs increases more rapidly with a than does @xmath387 , because of the increasing proportion of pairs with large separation . using the projection operators @xmath388 and @xmath389 for @xmath0 and 1 pairs we find that the total number of @xmath0 and 1 pairs in a nucleus depends only on its mass number @xmath32 and isospin @xmath390 : @xmath391 , \\ n_{1,0}^a+n_{1,1}^a &
= & \frac{1}{8}\left[3a^2 - 6a+4t_a(t_a+1)\right].\end{aligned}\ ] ] the above relations are obeyed by @xmath392 obtained from either the @xmath393 or correlated wave functions , since , in the present study , both are eigenstates of @xmath390 .
if the total spin , @xmath394 were to be a good quantum number we would have similar relations , @xmath395 , \\
n_{0,1}^a+n_{1,1}^a & = & \frac{1}{8}\left[3a^2 - 6a+4s_a(s_a+1)\right],\end{aligned}\ ] ] for the total number of pairs with spin 0 and 1 .
they are obeyed by the @xmath392 calculated for the @xmath393 states which have @xmath396 and 0 for @xmath4h , @xmath14he , @xmath15he , @xmath16li , @xmath259li and @xmath7o respectively .
however , @xmath397 is not a good quantum number ; tensor correlations admix states with larger @xmath397 in the ground state .
these reduce the @xmath398 and increase the @xmath399 by the same amount due to eq .
( 6.2 ) . in @xmath14he ( @xmath15he ) the @xmath399
is given by 1.5@xmath400 ( 3@xmath400 ) , where @xmath400 is the fraction of @xmath401 , @xmath402 ( @xmath403 ) state in the nuclear ground state .
the interaction in the @xmath21 state is much more attractive than that in the @xmath19 state .
hence the depletion of @xmath21 pairs by tensor correlations reduces the binding energy of nuclei significantly .
for example , in @xmath15he the @xmath21 interaction gives @xmath404 mev per pair , while the @xmath19 interaction gives only 0.8 mev per pair .
thus the conversion of 0.47 @xmath33=1 pairs from @xmath1=0 to @xmath1=1 state raises the energy of @xmath15he by @xmath405 mev . it should be stressed that this is a `` many - body '' effect absent in the two - body cluster approximation of either brueckner or variational methods .
the tensor interaction between nucleons @xmath135 and @xmath406 can flip their spins and convert pairs @xmath407 and/or @xmath408 from @xmath1=0 to @xmath1=1 .
the main conclusions of this study of nuclear structure , as predicted by realistic models of nuclear forces , are : \(i ) the static part of the two - nucleon potential in @xmath409 state has a large angular dependence due to the tensor interaction dominated by one - pion exchange . at @xmath410 fm
the difference between this potential at @xmath55 = @xmath58 and 0 is @xmath411 mev in most models ( fig .
[ potfig ] ) .
it confines @xmath409 pairs to the small @xmath55 region producing toroidal distributions .
the central hole in these tori is due to the repulsive core in @xmath69 interaction .
the maximum density in the tori is large , due to which the peak one - body density in deuterium exceeds 0.3 @xmath36 in most models .
\(iii ) the diameter of the maximum density torus , and the thickness of the half - maximum density torus are predicted to be @xmath413 and 0.9 fm , respectively ; these values are supported by the observed elastic electron - deuteron scattering .
\(v ) the pair distribution functions in @xmath20 and 1,0 states in different nuclei , can be scaled to lie on universal surfaces for @xmath414 fm .
these universal surfaces are predicted by the density distributions of the deuteron and the @xmath415 virtual bound state . the scaling factor @xmath235 for the @xmath20 densities provides a rigorous definition of the levinger - bethe quasi - deuteron number of the nucleus .
the calculated values of @xmath235 are significantly different from estimates based on independent - particle models , and in qualitative agreement with photon and pion absorption data .
\(vi ) the many - body distribution functions are also predicted to be anisotropic .
in particular the anisotropies of the @xmath329 , @xmath416 , and @xmath417 distributions in @xmath14he , @xmath15he and @xmath16li are strongly influenced by the toroidal structure of the deuteron .
\(vii ) tensor correlations convert @xmath33=1 pairs of nucleons from @xmath1=0 to @xmath1=1 state .
this many - body effect reduces the binding energies of nuclei
. it does not appear as if many - body effects reduce the magnitude of tensor correlations for the range of nuclei studied here : @xmath4h to @xmath7o .
due to the small size of this toroidal structure it may be worthwhile to attempt to understand it from the more basic quark degrees of freedom . within the constituent quark model @xcite
it requires a solution of the six - quark problem with a suitably chosen hamiltonian .
many attempts have been made ( see refs .
@xcite for example ) to calculate the nucleon - nucleon interaction from approximate solutions of the six - quark hamiltonian . a direct coupling of the pions to the quarks
is used to obtain the tensor part of the interaction .
the toroidal structure is presumably very sensitive to this coupling and to the tensor part of the quark - quark interaction in the framework of the constituent quark model .
as is well known , toroidal structure for the ground state of the deuteron was predicted many years ago @xcite using classical skyrme field theory @xcite related to qcd in the @xmath46 limit . in the classical limit one obtains a toroidal shape of @xmath418 fm in size and a binding energy of @xmath419 mev . from fig .
[ potfig ] it is obvious that in the classical limit realistic models of nuclear forces would also give a deuteron binding energy in the 100 to 200 mev range .
there have been attempts to include quantum corrections to this theory .
a recent calculation @xcite obtains an energy of 6.18 mev for the deuteron in this model .
ground states of the classical skyrme field with baryon numbers 3 to 6 have also been studied @xcite .
the baryon equidensity surfaces of these classical solutions are highly anisotropic
. however , the nucleon equidensity surfaces of the @xmath420 and @xmath421 @xmath14he and @xmath15he must be spherically symmetric , thus a direct comparison is not possible .
nevertheless the anisotropic @xmath329 and @xmath416 distributions in @xmath14he and @xmath15he may have some relation to the baryon density distributions in the skyrme model .
the authors wish to thank brian pudliner for many interesting discussions .
wishes to thank a. m. eir for useful discussions on the two - cluster overlap functions .
r.b.w . wishes to thank dieter kurath for useful comments on constructing wave functions for @xmath6li .
the calculations were made possible by grants of computer time from the mathematics and computer science division of argonne national laboratory , the pittsburgh supercomputing center , the cornell theory center and the national energy research supercomputer center .
the work of j.l.f , v.r.p . and a.a .
has been partially supported by u.s .
national science foundation via grant no .
phy94 - 21309 , that of s.c.p . and r.b.w .
by the u.s .
department of energy , nuclear physics division under contract no .
e-31 - 109-eng-38 , that of a.a .
by universidade de lisboa , junta de investigao cientfica e tecnolgica under contract no .
pbic / c / cen/1108/92 , and that of r.s . by the u.s .
department of energy .
electronic address : [email protected] electronic address : [email protected] electronic address : [email protected] electronic address : [email protected] electronic address : [email protected] electronic address : [email protected] r. g. arnold _ et al_. , phys .
lett . * 35 * , 776 ( 1975 ) .
g. g. simon , ch .
schmitt , and v. h. walther , nucl .
* a364 * , 285 ( 1981 ) .
r. cramer _
et al_. , z. phys .
c * 29 * , 513 ( 1985 ) .
s. platchkov _
et al_. , nucl .
phys . * a510 * , 740 ( 1990 ) . s. auffret _
et al_. , phys .
lett . * 54 * , 649 ( 1985 ) .
r. g. arnold _
et al_. , phys . rev .
* 58 * , 1723 ( 1987 ) .
m. e. schulze _
et al_. , phys .
* 52 * , 597 ( 1984 ) .
i. the _ et al_. , phys . rev .
* 67 * , 173 ( 1991 ) . v. f. dmitriev _ et al_. , phys .
* 157b * , 143 ( 1985 ) .
r. gilman _
et al_. , phys .
lett . * 65 * , 1733 ( 1990 ) .
w. fabian and h. arenhvel , nucl
* a258 * , 461 ( 1976 ) .
e. hummel and j. a. tjon , phys .
lett . * 63 * , 1788 ( 1989 ) .
r. schiavilla and d. o. riska , phys .
c * 43 * , 437 ( 1991 ) .
j. w. van orden , n. devine , and f. gross , phys .
lett . * 75 * , 4369 ( 1995 ) .
cdddddd @xmath422nucleus & @xmath235 & @xmath423 & @xmath424 & @xmath425 & + & & & & & ip & @xmath30 + @xmath14he & 2.0 & 2.1 & 2.4(1 ) & @xmath172 & 1.5 & 1.49 + @xmath15he & 4.7 & 5.1 & 4.3(6 ) & @xmath174 & 3 . &
2.99 + @xmath16li & 6.3 & 6.3 & & & 5.5 & 5.46 + @xmath259li & 7.2 & 7.8 & & 6.5(5 ) & 6.75 & 6.73 + @xmath7o & 18.8 & 22 & 17(3 ) & 16(3 ) & 30 . & 30.1 + cddddddddd @xmath422nucleus & @xmath426 & & @xmath427 & & @xmath379 & + & & ip & & & ip & & & ip & + @xmath14he & 0.087 & 1.5 & 1.35 & 0.0016 & 0 . & 0.01 & 0.012 & 0 . & 0.14 + @xmath15he & 0.22 & 3 . & 2.5 & 0.0085 & 0 . & 0.01 & 0.060 & 0 . & 0.47 + @xmath16li & 0.24 & 4.5 & 4.0 & 0.061 & 0.5 & 0.52 & 0.104 & 4.5 & 4.96 + @xmath259li & 0.37 & 6.75 & 6.1 & 0.118 & 0.75 & 0.77 & 0.18 & 6.75 & 7.41 + @xmath7o & 1 . & 30 . & 28.5 & 1 . & 6 . & 6.05 & 1 . & 54 . & 55.5 + | the two - nucleon density distributions in states with isospin @xmath0 , spin @xmath1=1 and projection @xmath2=0 and @xmath31 are studied in @xmath4h , @xmath5he , @xmath6li and @xmath7o .
the equidensity surfaces for @xmath2=0 distributions are found to be toroidal in shape , while those of @xmath2=@xmath31 have dumbbell shapes at large density .
the dumbbell shapes are generated by rotating tori .
the toroidal shapes indicate that the tensor correlations have near maximal strength at @xmath8 fm in all these nuclei .
they provide new insights and simple explanations of the structure and electromagnetic form factors of the deuteron , the quasi - deuteron model , and the @xmath9 , @xmath10 and @xmath11 @xmath12=2 ( @xmath13-wave ) components in @xmath14he , @xmath15he and @xmath16li .
the toroidal distribution has a maximum - density diameter of @xmath171 fm and a half - maximum density thickness of @xmath170.9 fm .
many realistic models of nuclear forces predict these values , which are supported by the observed electromagnetic form factors of the deuteron , and also predicted by classical skyrme effective lagrangians , related to qcd in the limit of infinite colors . due to the rather small size of this structure
, it could have a revealing relation to certain aspects of qcd .
experiments to probe this structure and its effects in nuclei are suggested .
pair distribution functions in other @xmath18 channels are also discussed ; those in @xmath19 have anisotropies expected from one - pion exchange interactions .
the tensor correlations in @xmath20 states are found to deplete the number of @xmath21 pairs in nuclei and cause a reduction in nuclear binding energies via many - body effects .
= 10000 |
the theory of derivations of various classes of banach @xmath0-algebras ( e.g. @xmath5 , @xmath6 and @xmath7-algebras ) is very well developed ( see , for example , @xcite ) .
it is well known that every derivation of a @xmath5-algebra is norm continuous and every derivation of a @xmath6-algebra ( in particular , of a @xmath7-algebra ) is inner @xcite .
the development of the theory of noncommutative integration , initiated by i. segal s paper @xcite prompted the introduction of numerous non - trivial @xmath0-algebras of unbounded operators , which , in a certain sense , are close to @xmath6 and @xmath7-algebras .
the main interest here is represented by the @xmath0-algebra @xmath1 ( respectively , @xmath8 ) of all locally measurable ( respectively , measurable ) operators , affiliated with a @xmath7-algebra ( or with a @xmath6-algebra ) @xmath2 and also by the @xmath0-algebra @xmath9 of all @xmath10-measurable operators from @xmath8 , where @xmath10 is a faithful normal semifinite trace on @xmath2 @xcite .
the importance of the algebra @xmath1 for the theory of unbounded derivations on von neumann algebras may be seen from the following classical example .
consider the algebra @xmath11 equipped with the semifinite trace given by lebesgue integration and consider ( a partially defined ) derivation @xmath12 on @xmath2 .
a simple argument shows that the algebra @xmath1 , which in this case coincides with the space of all measurable complex functions on @xmath13 is the only natural receptacle of @xmath14 .
similar examples can be produced in much more sophisticated circumstances and clearly indicate that the algebra @xmath1 is the most suitable object for studying unbounded derivations on a given von neumann algebra @xmath2 .
however , the study of derivations in the setting of @xmath1 has been greatly impeded by the fact that it is not a banach algebra ( it is not even a frechet algebra or locally convex algebra when endowed with its natural topology ) .
an additional difficulty ( especially , in comparison with rather well studied algebras @xmath9 ) is represented by the lack of developed analytical techniques in @xmath1 . only recently , in the series of papers @xcite meaningful attempts have been made to study the structure of derivations on such algebras .
of particular interest is the problem of identifying the class of von neumann algebras , for which any derivation of the @xmath0-algebra @xmath1 is inner . in the setting of commutative @xmath7-algebras ( respectively , commutative @xmath6-algebras )
this problem is fully resolved in @xcite ( respectively , in @xcite ) . in the setting of von neumann algebras of type @xmath15 , a thorough treatment of this problem
may be found in @xcite and @xcite .
the papers @xcite contain examples of non - inner derivations of the @xmath0-algebra @xmath1 , which are not continuous with respect to the topology @xmath3 of local convergence in measure on @xmath1 .
the latter topology is the only topology considered on algebras @xmath1 to date , it may be also viewed as a noncommutative generalization of the classical topology of convergence in measure on the sets of finite measure in the case when @xmath2 is given by the algebra @xmath16 , where @xmath17 is a @xmath18-finite measure space ( in this case the algebra @xmath1 coincides with the algebra of all measurable complex functions on @xmath19 ) .
see details in section 2 below . on the other hand ,
it is shown in @xcite that in the special case when @xmath2 is a properly infinite von neumann algebra of type @xmath15 , every derivation of @xmath1 is continuous with respect to the local measure topology @xmath3 . moreover , all such derivations are inner . using a completely different technique , a similar result was also obtained in @xcite under the additional assumption that the predual space @xmath20 to @xmath2 is separable .
it is of interest to observe that an analogue of this result ( that is the continuity of an arbitrary derivation of @xmath21 ) also holds for any von neumann algebra @xmath2 of type @xmath22 @xcite . in @xcite
the following problem is formulated ( problem 3 ) : let @xmath2 be a von neumann algebra of type @xmath23 and let @xmath10 be a faithful , normal , semifinite trace on @xmath2 .
is any derivation of a @xmath0-algebra @xmath9 equipped with the ( classical ) measure topology generated by the trace necessarily continuous ? in @xcite this problem is solved affirmatively for a properly infinite algebra @xmath2 . in view of the example we mentioned above , a natural problem ( analogous to problem 3 from @xcite )
is whether any derivation in a @xmath0-algebra @xmath1 is necessarily continuous with respect to the topology @xmath3 , where @xmath2 is a properly infinite von neumann algebra of type @xmath23 .
the main results of this paper provide an affirmative solution to this problem .
in fact , we establish a much stronger result that any derivation @xmath24 , where @xmath25 is any subalgebra in @xmath1 containing the algebra @xmath2 , is necessarily continuous with respect to the topology @xmath3 .
the proof proceeds in two stages .
firstly , we establish the @xmath3-continuity of any derivation @xmath26 for a properly infinite von neumann algebra @xmath2 ( section 3 ) .
then , in section 4 , a special construction of extension of a derivation @xmath27 up to a derivation defined on the whole algebra @xmath1 is given ( here @xmath2 is actually an arbitrary von neumann algebra ) .
we also hope that our approach to unbounded derivations on @xmath2 as well as techniques developed for dealing with locally measurable operators and the topology of local convergence in measure in this paper are of interest in their own right and may be used elsewhere .
we use terminology and notations from von neumann algebra theory @xcite and theory of locally measurable operators from @xcite .
let @xmath28 be a hilbert space , let @xmath29 be the @xmath0-algebra of all bounded linear operators on @xmath28 , and let @xmath30 be the identity operator on @xmath28 . given a von neumann algebra @xmath2 acting on @xmath28 , denote by @xmath31 the center of @xmath2 and by @xmath32 the lattice of all projections in @xmath2 .
let @xmath33 be the set of all finite projections in @xmath2 .
denote by @xmath34 the strong operator topology on @xmath29 , that is the locally convex topology generated by the family of seminorms @xmath35 , where @xmath36 is the hilbert norm on @xmath28 . a linear operator @xmath37 , where the domain @xmath38 of @xmath39 is a linear subspace of @xmath28 , is said to be _ affiliated _ with @xmath2 if @xmath40 for all @xmath41 from the commutant @xmath42 of algebra @xmath2 .
a densely - defined closed linear operator @xmath39 ( possibly unbounded ) affiliated with @xmath2 is said to be _
measurable _ with respect to @xmath2 if there exists a sequence @xmath43 such that @xmath44 and @xmath45 for every @xmath46 , where @xmath47 is the set of all natural numbers .
let us denote by @xmath8 the set of all measurable operators .
let @xmath48 .
it is well known that @xmath49 and @xmath50 are densely - defined and preclosed operators . moreover , the closures @xmath51 ( strong sum ) , @xmath52 ( strong product ) and @xmath50 are also measurable , and equipped with this operations ( see @xcite ) @xmath8 is a unital @xmath0-algebra over the field @xmath53 of complex numbers .
it is clear that @xmath2 is a @xmath0-subalgebra of @xmath8 .
a densely - defined linear operator @xmath39 affiliated with @xmath2 is called _ locally measurable _ with respect to @xmath2 if there is a sequence @xmath54 of central projections in @xmath2 such that @xmath55 and @xmath56 for all @xmath46 .
the set @xmath1 of all locally measurable operators ( with respect to @xmath2 ) is a unital @xmath0-algebra over the field @xmath53 with respect to the same algebraic operations as in @xmath8 @xcite and @xmath8 is a @xmath0-subalgebra of @xmath1 . if @xmath2 is finite , or if @xmath57 , the algebras @xmath8 and @xmath1 coincide ( * ? ? ?
* corollary 2.3.5 and theorem 2.3.16 ) .
if von neumann algebra @xmath2 is of type @xmath22 and @xmath58 , then @xmath59 and @xmath60 ( * ? ? ?
* theorem 2.2.19 , corollary 2.3.15 , ) . for every subset @xmath61 , the sets of all self - adjoint ( resp .
, positive ) operators in @xmath62 will be denoted by @xmath63 ( resp .
@xmath64 ) .
the partial order in @xmath1 is defined by its cone @xmath65 and is denoted by @xmath66 .
we shall need the following important property of the @xmath0-algebra @xmath1 .
let @xmath67 be a family of pairwise orthogonal non - zero central projections from @xmath2 with @xmath68 , where @xmath15 is an arbitrary set of indices ( in this case , the family @xmath67 is called a central decomposition of the unity @xmath30 ) .
consider the @xmath0-algebra @xmath69 with the coordinate - wise operations and involution and set @xmath70 @xcite,@xcite .
[ saito ] the mapping @xmath71 is a @xmath0-isomorphism from @xmath1 onto @xmath69 . observe that the analogue of proposition [ saito ] for the @xmath0-algebra @xmath8 does not hold in general @xcite .
proposition [ saito ] implies that given any central decomposition @xmath67 of the unity and any family of elements @xmath72 in @xmath1 , there exists a unique element @xmath73 such that @xmath74 for all @xmath75 .
this element is denoted by @xmath76 .
it is shown in @xcite that if @xmath2 is of type @xmath15 or @xmath22 , then for any @xmath77 there exists a countable central decomposition of unity @xmath54 , such that @xmath78 and @xmath79 for all @xmath80 .
let @xmath39 be a closed operator with dense domain @xmath81 in @xmath28 , let @xmath82 be the polar decomposition of the operator @xmath39 , where @xmath83 and @xmath84 is a partial isometry in @xmath29 such that @xmath85 is the right support @xmath86 of @xmath39 .
it is known that @xmath77 ( respectively , @xmath87 ) if and only if @xmath88 ( respectively , @xmath89 ) and @xmath90 @xcite . if @xmath39 is a self - adjoint operator affiliated with @xmath2 , then the spectral family of projections @xmath91 for @xmath39 belongs to @xmath2 @xcite . a locally measurable operator @xmath39 is measurable if and only if @xmath92 for some @xmath93 @xcite .
let us now recall the definition of the local measure topology .
first let @xmath2 be a commutative von neumann algebra .
then @xmath2 is @xmath0-isomorphic to the @xmath0-algebra @xmath16 of all essentially bounded measurable complex - valued functions defined on a measure space @xmath17 with the measure @xmath94 satisfying the direct sum property ( we identify functions that are equal almost everywhere ) ( see e.g. ( * ? ? ? * ch .
1 ) ) . the direct sum property of a measure @xmath94 means that the boolean algebra of all projections of the @xmath0-algebra @xmath16 is order complete , and for any non - zero @xmath95 there exists a non - zero projection @xmath96 such that @xmath97 .
the direct sum property of a measure @xmath94 is equivalent to the fact that the functional @xmath98 is a semi - finite normal faithful trace on the algebra @xmath99 . consider the @xmath0-algebra @xmath100 of all measurable almost everywhere finite complex - valued functions defined on @xmath17
( functions that are equal almost everywhere are identified ) .
on @xmath101 , define the local measure topology @xmath102 , that is , the hausdorff vector topology , whose base of neighborhoods of zero is given by @xmath103 @xmath104 where @xmath105 , @xmath106 , @xmath107 , and @xmath108 convergence of a net @xmath109 to @xmath110 in the topology @xmath102 , denoted by @xmath111 , means that @xmath112 in measure @xmath94 for any @xmath113 with @xmath107 .
note , that the topology @xmath102 does not change if the measure @xmath94 is replaced with an equivalent measure @xcite .
now let @xmath2 be an arbitrary von neumann algebra and let @xmath114 be a @xmath0-isomorphism from @xmath31 onto the @xmath0-algebra @xmath16 , where @xmath94 is a measure satisfying the direct sum property . denote by @xmath115 the set of all measurable real - valued functions defined on @xmath17 and taking values in the extended half - line @xmath116 $ ] ( functions that are equal almost everywhere are identified ) .
it was shown in @xcite that there exists a mapping @xmath117 that possesses the following properties : * @xmath118 ; * @xmath119 if @xmath120 ; * @xmath121 for any partial isometry @xmath90 ; * @xmath122 for any @xmath123 and @xmath124 ; * if @xmath125 , @xmath126 and @xmath127 , then @xmath128 .
a mapping @xmath129 that satisfies properties ( d1)(d5 ) is called a _ dimension function _ on @xmath130 .
a dimension function @xmath131 also has the following properties @xcite : * if @xmath132 , @xmath46 , then @xmath133 , in addition , when @xmath134 , @xmath135 , the equality holds ; * if @xmath136 , @xmath46 , @xmath137 , then @xmath138 almost everywhere . for arbitrary scalars
@xmath139 and a set @xmath140 , @xmath107 , we set @xmath141 @xmath142 where @xmath143 is the @xmath5-norm on @xmath2 .
it was shown in @xcite that the system of sets @xmath144 defines a hausdorff vector topology @xmath3 on @xmath1 such that the sets @xmath145 , @xmath105 , @xmath113 , @xmath107 form a neighborhood base of an operator @xmath77 .
it is known that @xmath146 is a complete topological @xmath0-algebra , and the topology @xmath3 does not depend on a choice of dimension function @xmath131 and on the choice of @xmath0-isomorphism @xmath114 ( see e.g. @xcite , @xcite ) . the topology @xmath3 on @xmath1 is called the _ local measure topology _ ( or the _ topology of convergence locally in measure _ ) .
note , that in case when @xmath147 the equality @xmath148 holds @xcite and the topology @xmath3 coincides with the uniform topology , generated by the @xmath5-norm @xmath149 .
we will need the following criterion for convergence of nets from @xmath1 with respect to this topology .
[ plm - spk1 ] ( i ) .
a net @xmath150 converges to zero with respect to the topology @xmath3 if and only if there is a net @xmath151 such that @xmath152 for all @xmath126 , @xmath153 , and @xmath154 , where @xmath102 is the local measure topology on @xmath155 , and @xmath114 is a @xmath0-isomorphism of @xmath31 onto @xmath16 .
a net @xmath156 converges to zero with respect to the topology @xmath3 if and only if @xmath157 for every @xmath93 , where @xmath158 is the spectral projection family for the operator @xmath159 .
[ rem_plm - spk1 ] it follows from proposition [ plm - spk1 ] that , if @xmath160 and @xmath161 , then @xmath162 .
since the involution is continuous in the topology @xmath3 , the set @xmath163 is closed in @xmath21 .
the cone @xmath65 of positive elements is also closed in @xmath21 @xcite .
proposition [ plm - spk1 ] will be used in the proof of the following convergence criterion .
[ p3 ] if @xmath164 , then @xmath165 fix a @xmath0-isomorphism @xmath166 and @xmath167 .
let @xmath168 be such that @xmath169 .
define the mapping @xmath170 by setting @xmath171 here , @xmath172 and @xmath173 is the restriction of @xmath94 to @xmath174 .
it is clear that @xmath175 is a @xmath0-isomorphism .
now define @xmath176 by setting @xmath177 for @xmath178 .
it is straightforward that @xmath179 is a dimension function on @xmath180 .
let @xmath181 .
we claim @xmath182 to see the claim , assume that the first convergence holds and observe that by proposition [ plm - spk1]@xmath183 , there exists a net @xmath184 such that @xmath185 for any @xmath186 , @xmath187 , and @xmath188 .
the projection @xmath189 belongs to the center @xmath190 of the von neumann algebra @xmath191 , and @xmath192 is a finite projection in @xmath191 for each @xmath126 .
also @xmath193 where @xmath194 is the local measure topology on @xmath195 , and @xmath196 hence , by proposition [ plm - spk1](i ) we get that @xmath197 .
we will show now that the convergence @xmath197 for @xmath181 implies the convergence @xmath198 .
let @xmath199 be a net in @xmath200 such that @xmath201 for every @xmath126 , @xmath202 and @xmath203 put @xmath204 . then @xmath205 and @xmath206 .
since @xmath207 , we have @xmath208 and @xmath209 also @xmath210 and so @xmath211 , since @xmath212 . again appealing to proposition [ plm - spk1](i ) ,
we conclude that @xmath198 .
now let @xmath213 and @xmath214 .
by proposition [ plm - spk1](ii ) , we have that @xmath215 for any @xmath93 , where @xmath158 is the spectral family for @xmath159 .
denote by @xmath216 the family of spectral projections for @xmath159 in @xmath217 , @xmath93 .
it is clear that @xmath218 and @xmath219 for all @xmath93 .
it follows from above that @xmath220 for all @xmath93 .
hence , by proposition [ plm - spk1](ii ) , it follows that @xmath221 .
the proof of the implication @xmath222 is similar and therefore omitted . the lattice @xmath130 is said to have a countable type , if every family of non - zero pairwise orthogonal projections in @xmath130 is , at most , countable .
a von neumann algebra is said to be @xmath18-finite , if the lattice @xmath223 has a countable type .
it is shown in ( * ? ? ?
* lemma 1.1 ) that a finite von neumann algebra @xmath2 is @xmath18-finite , provided that the lattice @xmath224 of central projections has a countable type .
if @xmath2 is a commutative von neumann algebra and @xmath130 has a countable type , then @xmath2 is @xmath0-isomorphic to a @xmath0-algebra @xmath16 with @xmath225 . in this case , the topology @xmath102 is metrizable and has a base of neighborhoods of @xmath226 consisting of the sets @xmath227 .
in addition , @xmath228 in measure @xmath94 , where @xmath229 .
let @xmath2 be a commutative von neumann algebra such that @xmath130 does not have a countable type .
denote by @xmath114 a @xmath0-isomorphism from @xmath2 on @xmath16 , where @xmath94 is a measure with the direct sum property .
due to the latter property , there exists a family @xmath67 of non - zero pairwise orthogonal projections from @xmath130 , such that @xmath230 and @xmath231 for all @xmath75 , in particular , @xmath232 has a countable type .
select @xmath233 so that @xmath234 and set @xmath235 let @xmath236 be the local measure topology on @xmath237 . since @xmath238 , we see that the topology @xmath236 coincides with the topology of convergence in measure @xmath239 in @xmath237 .
[ p4 ] for a net @xmath240 and @xmath110 from @xmath101 the following conditions are equivalent : ( i ) .
@xmath241 ; ( ii ) .
@xmath242 for all @xmath75 .
the implication @xmath243 follows from the definitions of topologies @xmath102 and @xmath236 .
it is sufficient to consider the case when @xmath245 .
consider the set @xmath246 of all finite subsets @xmath247 from @xmath15 and order it with respect to inclusion .
consider an increasing net @xmath248 in @xmath249 , where @xmath250 .
take an arbitrary neighborhood of zero @xmath251 ( in the topology @xmath102 ) and select @xmath252 in such a way that @xmath253 .
since @xmath254 , we can locate such @xmath255 that @xmath256 .
hence , @xmath257 for all @xmath126 . since @xmath258 for all @xmath259 and @xmath260 is a finite set , it follows @xmath261 .
thus , there exists such @xmath262 that @xmath263 for all @xmath264 .
in particular , @xmath265 which implies the convergence @xmath266 .
let us now establish a variant of proposition [ p4 ] for an arbitrary von neumann algebra @xmath2 .
let @xmath114 be a @xmath0-isomorphism from @xmath31 onto @xmath16 and let @xmath67 be a central decomposition of the unity . as before , we denote @xmath246 the directed set of all finite subsets from @xmath15 .
for every @xmath267 we set @xmath268 . since @xmath234 for some @xmath233 , we see that @xmath269 , where @xmath270 , and , in addition , @xmath271 which implies @xmath272 ( see proposition [ plm - spk1 ] ( i ) for @xmath273 ) .
as it the proof of proposition [ p4 ] for a given @xmath274 we choose @xmath255 such that @xmath275 for every @xmath77 .
if @xmath276 and @xmath277 for all @xmath75 , then by proposition [ p3 ] , we have @xmath278 for all @xmath75 , and so @xmath279 .
hence , there exists such @xmath262 that @xmath280 for all @xmath281 .
this means that @xmath282 the argument above justifies the following result .
[ p5 ] let @xmath2 be an arbitrary von neumann algebra , @xmath283 , @xmath284 when @xmath285 , @xmath286 .
the following conditions are equivalent : ( i ) .
@xmath287 ; ( ii ) .
@xmath288 for any @xmath75 .
[ r1 ] from propositions [ saito ] and [ p5 ] it follows that the topology @xmath3 coincides with the tikhonov product of topologies @xmath289 .
in particular , the isomorphism @xmath71 from proposition [ saito ] is a topological @xmath0-isomorphism from @xmath21 onto @xmath290 .
let @xmath67 be the same as in the assumption of proposition [ p5 ] , let @xmath291 be a linear operator such that @xmath292 for all @xmath293 .
it is clear that @xmath294 , @xmath295 is a linear operator acting in @xmath296 . due to proposition [ p5 ]
, the next result follows immediately .
[ c1 ] let @xmath2 and let @xmath67 satisfy the same assumptions of proposition [ p5 ] , and let @xmath297 be a linear operator such that @xmath292 for all @xmath298 .
the following conditions are equivalent : ( i ) .
the mapping @xmath299 is continuous ; ( ii ) .
the mapping @xmath300 is continuous for every @xmath75 .
let @xmath2 be an arbitrary von neumann algebra , let @xmath25 be a subalgebra in @xmath1 .
a linear mapping @xmath301 is called a _ derivation _ on @xmath25 with values in @xmath1 , if @xmath302 for all @xmath303 .
each element @xmath304 defines a derivation @xmath305 on @xmath25 with values in @xmath25 .
derivations @xmath306 are said to be _ inner derivations _ on @xmath25 .
since the operation of multiplication is continuous with respect to the topology @xmath3 , it immediately follows that any inner derivation of @xmath25 is continuous with respect to the topology @xmath3 .
now , we list a few properties of derivations on @xmath25 which we shall need below .
[ l1 ] if @xmath307 , @xmath14 is a derivation on @xmath25 and @xmath308 , then @xmath309 and @xmath310 for all @xmath311 .
we have that @xmath312 .
hence , @xmath313 , that is @xmath314 .
therefore , we have @xmath309
. in particular , @xmath315 .
let @xmath25 be an @xmath0-subalgebra in @xmath1 , let @xmath14 be a derivation on @xmath25 with values in @xmath1 .
let us define a mapping @xmath316 by setting @xmath317 , @xmath318 .
a direct verification shows that @xmath319 is also a derivation on @xmath25 .
a derivation @xmath14 on @xmath25 is said to be _ self - adjoint _ , if @xmath320 .
every derivation @xmath14 on @xmath25 can be represented in the form @xmath321 , where @xmath322 are self - adjoint derivations on @xmath25 .
since @xmath21 is a topological @xmath0-algebra , the following result holds .
[ l2 ] a derivation @xmath323 is continuous with respect to the topology @xmath3 if and only if the self - adjoint derivations @xmath324 and @xmath325 are continuous with respect to that topology .
as we already stated in the introduction , in the special case , when @xmath2 is a properly infinite von neumann algebra of type @xmath15 or von neumann algebra of type @xmath22 , any derivation of the algebra @xmath1 is continuous with respect to the topology @xmath3 @xcite .
the next theorem extends this result to an arbitrary properly infinite von neumann algebra .
[ main ] if @xmath2 properly infinite von neumann algebras , then any derivation @xmath326 is continuous with respect to the topology @xmath3 of local convergence in measure . by lemma [ l2 ] , we may assume that @xmath327 . since @xmath31 is a commutative von neumann algebra , there exists a system @xmath328 of non - zero pairwise orthogonal projections from @xmath31 such that @xmath68 and the boolean algebra @xmath232 has a countable type for all @xmath75 . by lemma [ l1 ]
we have that @xmath329 for all @xmath330 .
therefore , by corollary [ c1 ] , it is sufficient to prove that each derivation @xmath331 is @xmath332-continuous , @xmath75 .
thus , we may assume without loss of generality that the boolean algebra @xmath224 has a countable type . in this case
the topology @xmath3 is metrizable , and the sets @xmath333 , @xmath46 form a countable base of neighborhoods of @xmath226 ; in particular , @xmath21 is an @xmath334-space .
therefore it is sufficient to show that the graph of the linear operator @xmath14 is a closed set .
arguing by a contradiction , let us assume that the graph of @xmath14 is not closed .
this means that there exists a sequence @xmath335 , such that @xmath336 and @xmath337 . recalling that @xmath21 is a topological @xmath0-algebra and that @xmath320 , we may assume that @xmath338 for all @xmath46 . in this case ,
@xmath339 , where @xmath340 are respectively the positive and negative parts of @xmath39 . without loss of generality , we shall also assume that @xmath341 , otherwise , instead of the sequence @xmath342 we consider the sequence @xmath343 .
let us select scalars @xmath344 so that the projection @xmath345 does not vanish .
we have that @xmath346 and @xmath347 . replacing , if necessary , @xmath348 on @xmath349 , we may assume that @xmath350 by the assumption , @xmath2 is a properly infinite von neumann algebra and therefore , there exist pairwise orthogonal projections @xmath351 , such that @xmath352 and @xmath353 for all @xmath354 , in particular , @xmath355 . here , the notation @xmath356 denotes the equivalence of projections @xmath357 , and the notation @xmath358 means that there exists a projection @xmath359 such that @xmath360 .
in course of the proof of our main result we shall frequently use the following well - known fact : if @xmath356 and @xmath308 then @xmath361 . for every @xmath354 we select a projection @xmath362 , for which @xmath363 and denote by @xmath364 a partial isometry from @xmath2 such that @xmath365 .
clearly , we have @xmath366 whenever @xmath367 and the projection @xmath368 is infinite as the supremum of pairwise orthogonal and equivalent projections . taking into account that @xmath369 and @xmath370 we see that the series @xmath371 converges with respect to the topology @xmath34 to some operator @xmath372 satisfying @xmath373 in what follows
, we shall assume that the central support @xmath374 of the projection @xmath375 is equal to @xmath30 ( otherwise , we replace the algebra @xmath2 with the algebra @xmath376 ) .
let @xmath114 be a @xmath0-isomorphism from @xmath31 onto @xmath16 . by the assumption
, the boolean algebra @xmath224 has a countable type , and so we may assume that @xmath377 , where @xmath378 is the identity of the @xmath0-algebra @xmath16 . in this case ,
the countable base of neighborhoods of @xmath226 in the topology @xmath3 is formed by the sets @xmath379 .
recalling that we have @xmath336 and @xmath380 , we obtain @xmath381 when @xmath382 for every fixed @xmath354 .
fix @xmath383 , and using the convergence @xmath384 for @xmath385 , for each @xmath386 select an index @xmath387 and projections @xmath388 , @xmath389 , such that @xmath390 @xmath391 and @xmath392 for all @xmath393 .
similarly , using the convergence @xmath394 ( respectively , @xmath395 ) for @xmath396 , for each @xmath354 select indexes @xmath397 and @xmath398 and projections @xmath399 , @xmath400 , such that @xmath401 ( respectively , @xmath402 ) ; @xmath403 and @xmath404 , @xmath405 , for all @xmath406 ( respectively , @xmath407 ) . set @xmath408 , @xmath409 , @xmath410 . due to the selection of projections @xmath411 , @xmath412 and indexes
@xmath413 , we have that for each @xmath354 inequalities hold * @xmath414 ; * @xmath415 ; * @xmath416 ; * @xmath417 ; * @xmath418 + @xmath419 fix @xmath420 with @xmath421 and set @xmath422 since @xmath423 and @xmath424 , it follows that @xmath425 and @xmath426 , and therefore @xmath427 @xmath428 and @xmath429 inequalities ( [ m1m21])-([m1m23 ] ) mean that the sequence @xmath430 is a cauchy sequence in the @xmath334-space @xmath431 for each fixed @xmath383 .
consequently , there exists @xmath432 such that @xmath433 for @xmath434 , i.e. the series @xmath435 converges in @xmath1 with respect to the topology @xmath3 .
since the involution is continuous in topology @xmath3 and @xmath436 , we conclude @xmath437 . setting @xmath438 and using the relation @xmath439 ( see e.g. ( * ? ? ?
* ch . 5 , proposition 1.6 ) ) we have @xmath440 setting @xmath441 we have ( see ( [ p_0 ] ) , ( [ yp_0 ] ) and ( [ r_m ] ) ) @xmath442 from ( [ m1m21 ] ) it follows that @xmath443 since @xmath444 , and @xmath445 ( see ( [ r_m ] ) ) we obtain @xmath446 and hence , by ( [ dp_mq_m ] ) , @xmath447 due to ( [ r_m ] ) , we have @xmath448 for all @xmath386 .
hence , @xmath449 and @xmath450 using the properties of the derivation @xmath14 and equalities @xmath451 and ( [ r_m ] ) , ( [ q0z0 ] ) , we have @xmath452 consider the following formal series suggested by the preceding @xmath453 @xmath454 @xmath455 @xmath456 by the condition ( a2 ) the first series ( [ s1 ] ) and the second series ( [ s2 ] ) converge with respect to the norm @xmath457 to some elements @xmath458 respectively and @xmath459 and @xmath460 .
similarly , by the condition ( a3 ) , the third series ( [ s3 ] ) also converges with respect to the norm @xmath457 to some element @xmath461 , satisfying @xmath462 .
finally , since @xmath463 ( the convergence of the latter series is taken in the @xmath34 topology ) , we see that the fourth series ( [ s4 ] ) converges with respect to the topology @xmath34 to some element @xmath464 .
hence , the series @xmath465 converges with respect to the topology @xmath34 to some element @xmath466 , and , in addition , we have @xmath467 we shall show that @xmath468 where @xmath469 ( the convergence of the latter series is taken in the @xmath3-topology ( see ( [ y_k ] ) ) . using ( [ q0z0 ] ) for any @xmath420 we have @xmath470 since the series @xmath471 converges with respect to the topology @xmath34 ( see [ sq0deltaq0 ] ) , it follows that the series @xmath472 also converges with respect to this topology ( ( * ? ? ?
vi ) ) , in addition , the following equalities hold @xmath473 which guarantees @xmath474 since @xmath475 we see that for the right support @xmath476 of the operator @xmath477 satisfies the inequality @xmath478 and therefore @xmath479 consequently , @xmath480 for all @xmath481 . similarly , using the left support of the operator @xmath482
, we claim that @xmath483 . since @xmath484 , the equality ( [ qd ] ) holds .
thus , the inequality ( [ aq ] ) can be restated as follows @xmath485 it follows from the inequalities ( [ yq ] ) and ( [ aqk ] ) , that @xmath486 and @xmath487 due to ( [ yky ] ) , and taking into account ( [ p0q0 ] ) , we obtain @xmath488 that is @xmath489 let us now consider the projections @xmath490 using ( [ p0q0 ] ) , ( [ q0z01 ] ) we have that @xmath491 . therefore , combining ( [ dz_kp0q0k ] ) and ( [ q0z01 ] ) , we obtain @xmath492 that is the projection @xmath493 is finite ( see ( d1 ) ) . moreover , due to inequalities ( [ ky ] ) ( respectively , ( [ yqky ] ) ) , we have @xmath494 ( respectively , @xmath495 since @xmath114 is a @xmath0-isomorphism from @xmath496 onto @xmath16 , by ( [ z0 ] ) , we have that @xmath497 in particular , @xmath498 .
since @xmath499 and @xmath500 , we have @xmath501 , and therefore there exists such @xmath46 that @xmath502 ( see ( [ p_0 ] ) ) . since @xmath503 , we have @xmath504 for all @xmath354 .
hence , @xmath505 is an infinite projection . since the projection @xmath493 is finite ( see ( [ dq0z01 ] ) )
, we see that the projection @xmath506 must be infinite . by (
* proposition 6.3.7 ) , there exists a central projection @xmath507 such that @xmath508 is properly infinite , in particular , there exist pairwise orthogonal projections @xmath509 for all @xmath46 ( see , for example , ( * ? ? ? * proposition 2.2.4 ) ) .
in addition , @xmath510 for every @xmath46 the operator @xmath511 is locally measurable , and therefore there exists such a sequence @xmath512 that @xmath513 when @xmath514 and @xmath515 for all @xmath354 .
since @xmath516 it follows that @xmath517 when @xmath514 , and therefore , by ( [ int_e0q0 ] ) , for every @xmath46 there exists such a projection @xmath518 , that @xmath519 and @xmath520 consider the central projection @xmath521 since @xmath522 we have that @xmath523 for all @xmath46 .
due to ( [ int ] ) we have @xmath524 consequently , @xmath525 , and therefore @xmath526 from ( [ int_e0q0 ] ) and inequality ( [ phi ] ) , it follows that @xmath527 , i.e. @xmath528 and so @xmath529 . since @xmath508 is a properly infinite projection it follows that @xmath530 is a properly infinite projection . from the relationship @xmath531
, we see that the projection @xmath532 is also properly infinite for all @xmath46 . since @xmath533 it follows that @xmath534 is also properly infinite projection for every @xmath535 with @xmath536 . indeed , if @xmath537 and @xmath538 , then @xmath539 , and therefore , since the projection @xmath532 is properly infinite , we have @xmath540 .
consequently , the projection @xmath534 is also properly infinite . passing , if necessary to the algebra @xmath541
, we may assume that @xmath542 .
in this case , we also may assume that @xmath543 , @xmath544 and @xmath534 is a properly infinite projection for every non - zero @xmath308 .
the assumption @xmath545 means that for every fixed @xmath46 there exists such a sequence @xmath546 , that @xmath547 when @xmath514 and @xmath548 for all @xmath354 . since @xmath549 and @xmath550 ( see ( d7 ) )
, it follows that @xmath551 converges in measure @xmath94 to zero .
consequently , we may select a central projection @xmath552 and a finite projection @xmath553 as to guarantee @xmath554 , @xmath555 and @xmath556 for all @xmath46 . setting @xmath557
we have that @xmath558 this means that @xmath559 and @xmath560 ( see ( d1 ) ) .
in addition , since @xmath561 from ( [ f_nb_n ] ) it follows that @xmath562 for all @xmath46 .
consider the projections @xmath563 and @xmath564 .
clearly ( see ( [ e0 ] ) ) , @xmath565 for all @xmath46 , and also @xmath566 that is @xmath567 .
hence , for every non - zero central projection @xmath568 , we have that the projection @xmath569 is finite .
since the projection @xmath534 is infinite , the projection @xmath570 is also infinite , i.e. @xmath571 for any @xmath572 and @xmath46 . since @xmath573
, we see that there exists such an increasing sequence @xmath574 that @xmath575 for all @xmath46 . appealing to the inequalities ( [ kyq ] ) , ( [ gze ] ) and taking into account the equality @xmath576
, we deduce @xmath577 hence , @xmath578 for every @xmath579 the inequalities @xmath580 holds , in particular , @xmath581 .
hence , inequality ( [ ineq_l_n - n ] ) implies that @xmath582 since @xmath583 whenever @xmath135 , we see ( due to inequalities ( [ kyq ] ) and ( [ gze ] ) ) that the series @xmath584 converges with respect to the topology @xmath34 to a self - adjoint operator @xmath585 , satisfying @xmath586 again appealing to the inequalities ( [ kqy ] ) , ( [ gze ] ) and ( [ ge ] ) , we infer that @xmath587 thus , @xmath588 for every @xmath46 .
set @xmath589 , where @xmath590 is the spectral family of projections for self - adjoint operator @xmath591 . for every @xmath46 we have @xmath592 hence , @xmath593 which implies @xmath594 i.e. @xmath595 .
then @xmath596 , and therefore @xmath597 for all @xmath46 .
since @xmath598 , we see that there exists such a non - zero central projection @xmath599 , that @xmath600 .
hence , we may find such @xmath601 , that @xmath602 for all @xmath603 ( @xcite ) , that is @xmath604 when @xmath605 . since @xmath606
, we infer from ( [ dgn ] ) that @xmath607 for all @xmath608 which contradicts with the property ( d1 ) in the definition of the dimension function @xmath131 , since @xmath609 is an infinite projection ( see ( [ zgn ] ) ) .
hence , our assumption that the derivation @xmath14 fails to be continuous in @xmath21 has led to a contradiction .
observe that in the special case of properly infinite von neumann algebras of type @xmath15 or @xmath22 , theorem [ main ] gives a new proof of the results concerning the continuity of a derivation of @xmath21 established earlier in @xcite .
in this section the construction of extension of any derivation , acting on a von neumann algebra @xmath2 with values in @xmath1 , up to a derivation from @xmath1 into @xmath1 is given . using this extension and theorem [ main ]
it is established that in case the of a properly infinite von neumann algebra @xmath2 , any derivation @xmath24 from a subalgebra @xmath25 satisfying @xmath611 is continuous with respect to the local measure topology .
let @xmath612 be an arbitrary von neumann algebra and let @xmath54 be a sequence of central projections from @xmath612 , such that @xmath613 .
a sequence @xmath614 is called _ consistent _ with the sequence @xmath54 , if for any @xmath615 the equality @xmath616 holds for @xmath617 .
[ p8 ] let @xmath618 ( respectively , @xmath619 ) be a sequence consistent with the sequence @xmath620 ( respectively , with the sequence @xmath621 ) , @xmath613 ( @xmath622 ) . then ( i ) .
there exists a unique @xmath77 , such that @xmath623 for all @xmath46 , in addition , @xmath624 ; ( ii ) . if @xmath625 for all @xmath626 , then @xmath627 for @xmath382 .
consider a neighborhood @xmath274 of zero in topology @xmath3 , where @xmath628 ( see the definition of topology @xmath3 in section 2 ) .
since @xmath629 , it follows that @xmath630 for @xmath631 .
taking @xmath632 , we have @xmath633 , i.e. @xmath634 for all @xmath635 . for @xmath636
, we have @xmath637 for all @xmath631 .
it means that @xmath638 is a cauchy sequence in @xmath21 .
consequently , there exists @xmath77 such that @xmath639 . since @xmath640 for all @xmath631
, it follows that @xmath641 , and therefore @xmath642 .
fixing @xmath383 , for @xmath643 we have @xmath644 for @xmath382 , i.e. @xmath645 for all @xmath383 . if @xmath646 and @xmath647 for all @xmath46 , then @xmath648 , i.e. @xmath649 .
( ii ) . if @xmath650 for @xmath382 , @xmath651 for @xmath514 , and @xmath652 for @xmath653 , then @xmath654 for @xmath653 .
now , we consider a derivation @xmath14 from @xmath8 into @xmath1 and construct an extension @xmath655 from @xmath1 into @xmath1 .
recall that for an arbitrary operator @xmath77 there exists a sequence @xmath656 such that @xmath657 and @xmath658 for all @xmath46 . since @xmath659 ( see lemma [ l1 ] ) , the sequence @xmath660 is consistent with the sequence @xmath54 . by proposition [ p8](i ) , there exists a unique @xmath661 such that @xmath662 ( notation : @xmath663 ) . set @xmath664 . according to proposition [ p8](ii ) , the definition of operator @xmath665 does not depend on a choice of a sequence @xmath620 , for which @xmath657 and @xmath658 , @xmath46 .
if @xmath87 , then , taking @xmath666 , we obtain @xmath667 .
[ p9 ] the mapping @xmath655 is a unique derivation from @xmath1 into @xmath1 such that @xmath667 for all @xmath87 .
let @xmath668 , and let @xmath669 be such that@xmath670 . observing that @xmath671 we have @xmath672 similarly , @xmath673 , @xmath674 .
further , using convergences @xmath675 and the inclusion @xmath676 , we have @xmath677 consequently , @xmath678 is a derivation , in addition , @xmath667 for all @xmath87 .
assume that @xmath679 is also a derivation for which @xmath680 for all @xmath87 .
let us show that @xmath681 . if @xmath682 , then , by lemma [ l1 ] and proposition [ p8 ] ( i ) , we obtain @xmath683 now , we give the construction of extension of a derivation @xmath684 up to a derivation @xmath685 . for each @xmath77 set @xmath686 , where @xmath687 is the left and @xmath86 is the right support of @xmath39 . if @xmath82 is a polar decomposition of @xmath688 , then @xmath689 @xcite and , due to equalities @xmath690 , we have @xmath691 .
we need the following lemma .
[ l3 ] if @xmath131 is a dimension function of a von neumann algebra @xmath2 , then for any derivation @xmath14 from @xmath2 into @xmath1 the following inequality @xmath692 holds for all @xmath693 .
for @xmath693 we have @xmath694 i.e. @xmath695 and @xmath696 that implies the inequalities ( see ( d2 ) , ( d3 ) ) @xmath697 since @xmath698 we have @xmath699 due to ( d6 ) , we have @xmath700 as in the definition of the topology @xmath3 , denote by @xmath114 a @xmath0-isomorphism from @xmath31 onto the @xmath0-algebra @xmath16 , where @xmath94 is a measure satisfying the direct sum property . by proposition [ plm - spk1](i ) ,
the convergence of the sequence of projections @xmath701 is equivalent to existence of a sequence @xmath702 such that @xmath703 for all @xmath704 , @xmath705 and @xmath706 .
[ l4 ] if @xmath707 , then @xmath336 .
taking @xmath708 for all @xmath46 , we have @xmath709 and @xmath710 consequently , @xmath711 ( see proposition [ plm - spk1](i ) ) . since @xmath712 for all @xmath713 , it follows @xmath714 , and therefore @xmath336 ( see remark [ rem_plm - spk1 ] ) . if @xmath715 and @xmath137 , then @xmath716 ( see ( d1 ) ) and @xmath717 ( see ( d2 ) and d(7 ) ) , in particular , @xmath718 .
hence , lemma [ l4 ] implies the following [ cc1 ] if @xmath719 , then @xmath701 .
[ l6 ] let @xmath720 .
if @xmath684 is a derivation , then there exists @xmath721 , such that @xmath722 for @xmath617 we have @xmath723 and therefore , applying lemma [ l3 ] and properties ( d2 ) , ( d3 ) , we obtain @xmath724 since @xmath725 ( see ( d1 ) ) and @xmath726 ( see ( d7 ) ) it follows that @xmath727 ( see ( d7 ) ) .
hence , @xmath728 for @xmath653 . by lemma [ l4 ]
, we have that @xmath729 for @xmath653 , i.e. @xmath730 is a cauchy sequence in @xmath21 .
consequently , there exists @xmath721 , such that @xmath731 let us show that @xmath732 . for each @xmath46
we have @xmath733 hence , @xmath734 since @xmath735 and @xmath736 it follows @xmath737 since ( see lemma [ l3 ] ) @xmath738
we have @xmath739 by lemma [ l4 ] , we obtain @xmath740 now , equipped with lemma [ l6 ] , we may extend any derivation @xmath684 up to a derivation @xmath741 from @xmath8 into @xmath1 .
for each @xmath87 there exists a sequence @xmath742 , such that @xmath743 for all @xmath46 . by lemma [ l6 ] , there exists @xmath721 , such that @xmath744 .
in addition , the definition of @xmath745 does not depend on a choice of sequence @xmath746 satisfying the above mentioned property , in particular , @xmath747 for all @xmath693 ( in this case , @xmath748 ) .
[ p10 ] the mapping @xmath741 is a unique derivation from @xmath8 into @xmath1 , such that @xmath747 for all @xmath693 . for @xmath48 select @xmath749 , such that @xmath750 for all @xmath46 .
the sequence of projections @xmath751 is increasing , and , in addition , @xmath752 the last estimate implies the convergence @xmath753 ( see ( d7 ) ) , or @xmath754 .
by lemma [ l6 ] , we have @xmath755 similarly , @xmath756 for all @xmath674 .
let us show that @xmath757 .
due to polar decomposition @xmath758 , we have @xmath759 for all @xmath46 . set @xmath760 since @xmath761
we obtain @xmath762 since @xmath87 , there exists @xmath763 such that @xmath764 for all @xmath765 , and therefore @xmath766 for all @xmath767 .
the equality @xmath768 implies that @xmath769 in particular , @xmath770 for all @xmath46 . from here and from the inequalities
@xmath771 it follows that @xmath772 . since @xmath773
, we have @xmath774 for @xmath775 for some @xmath776 .
hence , @xmath777 for @xmath775 and @xmath778 i.e. @xmath779 or @xmath780 .
using corollary [ cc1 ] , lemma [ l6 ] , the inclusions @xmath781 and equalities @xmath782 we obtain @xmath783 consequently , @xmath784 is a derivation , such that @xmath747 for all @xmath693 .
let @xmath785 also be a derivation , for which @xmath680 for all @xmath693 . if @xmath87 , then @xmath786 for all @xmath787 for some @xmath788 .
hence , @xmath789 ( see corollary [ cc1 ] ) .
since @xmath21 is a topological algebra , it follows that @xmath790 since @xmath791 for @xmath77 , it follows via lemma [ l3 ] , that @xmath792 by lemma [ l4 ] , we obtain @xmath793 , that implies the equality @xmath794 .
propositions [ p9 ] and [ p10 ] imply the following theorem , which is the main result of this section . [ ext ]
let @xmath25 be a subalgebra of @xmath1 , @xmath795 and let @xmath796 be a derivation .
then there exists a unique derivation @xmath797 such that @xmath798 for all @xmath799 . since @xmath795 ,
the restriction @xmath800 of the derivation @xmath14 on @xmath2 is a well - defined derivation from @xmath2 into @xmath1 .
hence , by propositions [ p9 ] and [ p10 ] , the mapping @xmath801 is a unique derivation from @xmath1 into @xmath1 such that @xmath802 for all @xmath693 .
let us show that @xmath803 for every @xmath804 .
if @xmath804 , then there exists a sequence @xmath620 , such that @xmath657 and @xmath805 . since @xmath806
( see proposition [ p8](i ) ) , we have , by lemma [ l1 ] , @xmath807 and , similarly , @xmath808 . using the equality @xmath809 for each @xmath693 , and following the proof of uniqueness of the derivation @xmath741 from proposition [ p10 ]
, we obtain @xmath810 for all @xmath46 , that implies the equality @xmath803 .
the following corollary immediately follows from theorems [ main ] and [ ext ] .
[ c2 ] let @xmath2 be a properly infinite von neumann algebra , @xmath25 is a subalgebra in @xmath1 and @xmath795 .
then any derivation @xmath796 is continuous with respect to the local measure topology @xmath3 .
in particular , corollary [ c2 ] implies that for a properly infinite von neumann algebra @xmath2 any derivation @xmath811 is @xmath3-continuous .
note , that in case , when @xmath2 is of type @xmath812 , any derivation of @xmath8 is inner @xcite , and therefore is automatically continuous with respect to the topology @xmath3 .
let @xmath2 be a semifinite von neumann algebra acting on hilbert space @xmath28 , @xmath10 be a faithful normal semifinite trace on @xmath2 .
an operator @xmath87 with domain @xmath81 is called _ @xmath10-measurable _ if for any @xmath813 there exists a projection @xmath814 such that @xmath815 and @xmath816 .
the set @xmath9 of all @xmath10-measurable operators is a @xmath0-subalgebra of @xmath8 such that @xmath817 .
if the trace @xmath10 is finite , then @xmath818 .
the algebra @xmath9 is a noncommutative version of the algebra of all measurable complex functions @xmath110 defined on @xmath17 , for which @xmath819 for @xmath820 . for each @xmath821
it is possible to define the generalized singular value function @xmath822 which allows to define and study a noncommutative version of rearrangement invariant function spaces .
for the theory of the latter spaces , we refer to @xcite,@xcite .
the pair @xmath826 is a complete metrizable topological @xmath0-algebra . here , the topology @xmath823 majorizes the topology @xmath3 on @xmath9 and , if @xmath10 is a finite trace , the topologies @xmath823 and @xmath3 coincide @xcite .
however , if @xmath827 , then on @xmath826 topologies @xmath823 and @xmath3 do not coincide in general @xcite . for example , when @xmath828 where @xmath94 is a @xmath18-finite measure , @xmath829 , the topology @xmath823 in @xmath830 coincide with the topology of convergence in measure @xmath94 , and the topology @xmath102 is the topology of convergence locally in measure @xmath94 ( section 2 ) , in particular , if @xmath831 and @xmath832 , then @xmath833 , whereas @xmath834 .
see the detailed comparison of topologies @xmath823 and @xmath3 in @xcite .
it is proved in @xcite that in a properly infinite von neumann algebra @xmath2 any derivation @xmath835 is continuous with respect to the topology @xmath823 .
corollary [ c2 ] implies that , in this case , derivation @xmath835 is continuous with respect to the topology @xmath3 too .
recall @xcite , that a @xmath0-subalgebra @xmath25 of @xmath1 is called absolutely solid if conditions @xmath836 imply that @xmath799 . in ( * ? ?
* proposition 5.13 ) it is proved that , if @xmath14 is a derivation on absolutely solid @xmath0-subalgebra @xmath837 and @xmath838 $ ] for all @xmath799 and some @xmath839 , then there exists @xmath840 , such that @xmath841 $ ] for all @xmath799 , i.e. the derivation @xmath14 is inner on the @xmath0-subalgebra @xmath25 . this observation and theorem [ ext ] yield our final result [ c3 ] suppose that all derivations on the algebra @xmath1 are inner and let @xmath837 be an absolutely solid @xmath0-subalgebra of @xmath1 .
then all derivations on @xmath25 are inner .
in particular , any derivation on the algebras @xmath8 and @xmath9 are inner .
the result of corollary [ c3 ] extends and generalizes ( * ? ? ?
* theorem 6.8 ) and ( * ? ? ?
* proposition 5.17 ) .
note , that the conditions of corollary [ c3 ] hold , in particular , for properly infinite von neumann algebras , which do not have direct summand of type @xmath842 @xcite,@xcite,@xcite .
a.f.ber , _ continuity of derivations on properly infinite * -algebras of @xmath10-measurable operators _ , mat .
zametki * 90 * ( 2011 ) no . 5 , 776780 ( russian ) .
english translation : _ math . notes _ , * 90 * ( 2011 ) no . 5 - 6 , 758762 . | we prove that any derivation of the @xmath0-algebra @xmath1 of all locally measurable operators affiliated with a properly infinite von neumann algebra @xmath2 is continuous with respect to the local measure topology @xmath3 .
building an extension of a derivation @xmath4 up to a derivation from @xmath1 into @xmath1 , it is further established that any derivation from @xmath2 into @xmath1 is @xmath3-continuous . |
it can be regarded as one of the most intriguing results of half a century in physics of hot magnetized plasmas , that in macroscopically stable equilibria the confinement is nevertheless determined by micro - scale instabilities and fluctuations , with transport of energy and particles across magnetic flux surfaces being dominated by ubiquitous turbulence on small ( gyro radius ) scales and low ( drift ) frequencies @xcite .
this turbulence is regulated by the formation of mesoscopic zonal structures out of the turbulent flows @xcite .
criteria for design of new magnetized plasma experiments for fusion research like iter are primarily based on empirical scaling laws for the energy confinement time @xcite .
first - principle based transport models still require reference to experimental scalings , and are validated mainly only for core plasmas excepting the pedestal region @xcite .
an improved understanding of the underlying turbulent dynamics and of its relation to design parameters , like the shape of the confining magnetic field , will facilitate the development of advanced magnetic confinement experiments . the plasma shape of a tokamak enters into confinement and transport modelling through parameters specifying a vertical elongation @xmath0 and an outboard side triangularity @xmath1 that describe the deviation from a simple circular torus .
the design criteria of the iter plasma shape with @xmath2 and @xmath3 ( at @xmath4 flux surface ) are based on a conservative regime that is well established in present experiments @xcite .
more extreme variations of flux surface shaping were investigated with the experiment tcv ( tokamak configuration variable ) which is able to achieve configurations up to @xmath5 and @xmath6 @xcite .
experimental evidence suggests that large elongation is always beneficial for local and global energy confinement , whereas the role of triangularity depends also on the value of the plasma pressure gradient .
poloidal cross section of a tokamak plasma ( asdex upgrade ) showing the elongation and triangularity of closed flux surfaces and a lower x - point on the separatrix .
a section of the flux tube for drift wave turbulence computations winding around the torus along field lines is here locally intersecting the outer midplane of the plasma edge.,width=321 ] an important task for computational plasma physics is to determine the effects of flux surface shaping on turbulence and transport by first - principle direct numerical simulation .
shaping effects are complicated even for apparently simple circular equilibria @xcite , and efforts to perform modelling for edge plasmas most often incorporate technically complicated prescriptions in an inflexible way @xcite .
an advantageous way to explore shaping is through the use of a covariant metric in a model with cartesian topology @xcite , but even then it was found that the use of experimental geometry , while most realistic for modelling purposes , may not be helpful to theoretical understanding , and hence standard `` kappa - delta '' models were used @xcite . in attempt to understand the effects of stellarator geometry , a set of analytical forms were used for the metric , varying each in turn to provide a small set of steps from tokamak to stellarator geometry @xcite .
forms like this were also used to characterise the x - point effects , again making simple control tests tractable @xcite . in the following ,
we report on the results of a more systematic series of `` kappa - delta '' tokamak configurations with varying elongation and triangularity , and an actual asdex upgrade @xcite divertor configuration with the equilibrium code helena @xcite .
these are implemented into the gyrofluid code gem ( @xcite , which is updated from ref .
@xcite ) for computations of edge turbulence and flows .
specific effects of flux surface shaping on drift - alfvn ( dalf ) and ion temperature gradient ( itg ) turbulence in the tokamak edge plasma are analysed and compared , and the mechanisms of the coupling between geometry and turbulence are identified .
the emphasis is on physical mechanism , which is complimentary to and should also be helpful to the understanding of the more experimentally oriented modelling efforts .
it is found that turbulent transport fluxes are reduced by increasing flux surface elongation @xmath7 .
the influence of triangularity @xmath8 is much weaker and generally also depends on elongation and on a pressure shift .
the geodesic acoustic oscillation modes ( gam ) and the overall turbulence frequency spectra depend on flux surface geometry .
an enhancement of zonal flow amplitude in the spectra by elongation and x - point shaping is found to be weak in the dalf case and more pronounced in the itg case .
the reduction of transport observed in our simulations in shaped tokamak geometry is mainly a result of local and global magnetic shear .
we begin in section ii with a brief review and discussion of the role of sheared flows in tokamak turbulence in context of the toroidal flux surface geometry . in section iii the gyrofluid electromagnetic model gem for edge turbulence computations is presented , and the flux tube representation of tokamak geometry within a gem turbulence code is discussed in section iv .
transport scalings by elongation and triangularity , that are obtained with this model , are presented in section v. resulting frequency and wave number spectra in tokamak divertor geometry are discussed in section vi .
conclusions and implications of the results are presented in section vii .
one of the main requirements for successful performance of fusion experiments is the prospect of operation in a high confinement h - mode @xcite . the formation of a characteristical transport barrier in magnetically confined toroidal h - mode plasmas is closely related to the presence of a radially varying electric field near the last closed flux surface @xcite . a radial electric field , @xmath9 , gives rise to a perpendicular e - cross - b ( exb ) flow through the relation @xmath10 .
if @xmath9 varies with @xmath11 , then this flow is sheared , and this flow shear is what is invoked to explain the transport barrier @xcite .
the flow shear ( vorticity ) is expected to suppress turbulent transport either by shearing the eddies apart @xcite or by simply tilting them so that their reynolds - stress interaction with the vorticity layer tends towards energy transfer out of the eddies and into the flow @xcite .
neoclassical equilibrium effects can also produce these shear layers @xcite . as opposed to these mean exb flows , the fluctuating zonal flows in direct interaction with the turbulence
have also been observed in tokamak edge regions @xcite .
the transition between states of low ( l ) plasma confinement to the high ( h ) confinement regime featuring such an edge transport barrier is up to now not satisfactorily described by any of the existing first - principle theories @xcite . the notion that generation of zonal flows by direct drive of the turbulence can act to self - regulate by suppressing the driving turbulent vortices has lead to the development of predator - prey type bifurcation models that are able to describe specific characteristics of the l - h transition ( see e.g. @xcite ) .
early computations based on the resistive - g @xcite and collisional drift wave @xcite models found self - generated sheared flows which could be argued as an l - h transition trigger .
however , more recent models in more comprehensive treatment of the toroidal geometry and also the full range of scales of motion do not find this @xcite .
the phenomenology of these self consistent zonal flows receives wide interest @xcite , but they serve as a moderator of the turbulence rather than a suppression mechanism , as the energetic interactions with not only the pressure but also magnetic parts of the equilibrium tend to place each part of the overall dynamical system , which naturally includes the flows themselves @xcite . after the initial discovery of the h - mode in the divertor experiment asdex @xcite it has been found in numerous large tokamaks around the world that the presence of a divertor ( originally designed to improve the impurity exhaust ) significantly enhances the prospect of reaching this state @xcite . however , up to now both the transition models as well as the numerical simulations have relied mainly on simplified representations of the plasma geometry . in our present work
we investigate the effects of realistic flux surface shaping on tokamak edge turbulence using a gyrofluid model recently corrected for energy conservation properties @xcite , thereby updating previous gyrofluid results @xcite .
this provides a counterweight to existing efforts based on collisional fluid equations and associated numerical schemes which can not treat the ion gyroradius scales and therefore overestimate linear phenomena based in the larger scales @xcite .
the ion gyroradius necessarily enters because the dynamics is sensitive to phenomena in the range of the spectrum with perpendicular wavenumbers only slightly less than the inverse of the drift scale ( @xmath12 , also called `` ion sound gyroradius '' ) , and with comparable ion and electron temperatures @xmath12 is commensurate with the ion gyroradius @xmath13 @xcite
. a schematic view of a shaped plasma cross section including a flux - tube section from turbulence computations is shown in fig .
[ f : aug ] .
transport in a fusion edge plasma is dominated by turbulent low - frequency drift wave motion that causes a fluid - like convection through exb vortices in a plane perpendicular to the magnetic field direction , acting in concert with parallel coupling dynamics tending towards dissipation @xcite . in general
this is both nonlinear and electromagnetic @xcite .
toroidal compressibility of zonal ( flux surface averaged ) exb flows opens channels in the nonlinear dynamics of the vortex - flow interaction that strongly affect the flow energetics @xcite .
the origin of the compression is the geodesic curvature of the magnetic field lines .
the flow is thereby coupled with poloidally asymmetric pressure sidebands which are consumed by the turbulence and global alfvn or parallel flow dynamics .
this energetics was given in ref .
@xcite and the various coupling effects throughout the flow / equilibrium system were analysed in refs .
@xcite , and the sensitivity of the turbulence to the geodesic energy transfer effect was explored using an additional scaling parameter for the strength of its coupling @xcite .
this geodesic transfer mechanism represents a restoring loss channel for the zonal flows , ultimately placing them in statistical equilibrium with the turbulence , with the reynolds stress ( spin - up ) mechanism continuing to operate . since the latter acts on all frequencies , the geodesic acoustic oscillation ( gam ) @xcite itself need not be present for the geodesic transfer mechanism to operate .
the net result of this is that turbulent transport in 3d toroidal edge computations is found to be reduced but not completely suppressed by the self - generated zonal flows .
this geodesic transfer effect was first studied for the case of a simple circular toroidal magnetic field @xcite , in the following we give both numerical evidence and a physical picture of how plasma shaping and the presence of a divertor via its modification of plasma geometry can influence the outcome of the gyrofluid turbulence / flow interaction .
the dynamical character of cross - field exb drift wave turbulence in the edge region of a tokamak plasma is governed by electromagnetic and dissipative effects in the parallel response .
the most basic drift - alfvn ( dalf ) model to capture the drift wave dynamics includes nonlinear evolution equations of three fluctuating fields : the electrostatic potential @xmath14 , electromagnetic potential @xmath15 , and density @xmath16 .
the tokamak edge usually features a more or less pronounced density pedestal , and the dominant contribution to the free energy drive to the turbulence by the inhomogeneous pressure background is thus due to the density gradient . on the other hand , a steep enough ion temperature gradient ( itg ) in the edge does not only change the turbulent transport quantitatively , but adds new interchange physics into the dynamics .
in addition , more field quantities have to be treated : parallel and perpendicular temperatures @xmath17 and @xmath18 and the associated parallel heat fluxes , for a total of six moment variables for each species .
finite larmor radius effects introduced by warm ions require a gyrofluid description of the turbulence equations .
both the resistive dalf and the itg models can be covered by using the six - moment electromagnetic gyrofluid model gem @xcite , but we will refer for the dalf model to its more economical two - moment version @xcite .
the gyrofluid model is based upon a moment approximation of the underlying gyrokinetic equation .
the first complete six - moment formulation was given for slab geometry @xcite , and later extended to incorporate toroidal effects @xcite using a ballooning - based form of flux surface geometry @xcite .
electromagnetic induction and electron collisionality were then included to form a more general gyrofluid for edge turbulence @xcite , with the geometry correspondingly replaced by the version from the edge turbulence work , which does not make ballooning assumptions and in particular represents slab and toroidal eigenmode types equally well and does not require radial periodicity @xcite .
energy conservation considerations were solidified first for the two - moment version @xcite , and most recently for the six - moment version @xcite .
this latter reference currently defines the gem model , energetics , and also the numerical techniques .
the present paper uses the six - moment equation set for each species ( eqs .
99 - 104 ) , induction and polarisation equations ( eqs .
89,92 ) , the zero gyroradius limit for electrons , the local model with separate linear drive terms and dirichlet boundary conditions ( sec .
iv a ) , pad forms for gyroaveraging operators ( sec .
iv b ) , and the dissipation model and numerical scheme ( secs .
iv c - e ) , all with equation and section numbers from ref . @xcite .
this is the standard local form for the gem code . the flux surface geometry of a tokamak enters into the gyrofluid equations via the curvilinear generalisation of differentiation operators and via inhomogeneity of the magnetic field strength @xmath19 .
the different scales of equilibrium and fluctuations parallel and perpendicular to the magnetic field motivate the use of field aligned flux coordinates .
the differential operators in the field aligned frame are the parallel gradient @xmath20 with magnetic field disturbances @xmath21 as additional nonlinearities , the perpendicular laplacian @xmath22,\ ] ] and the curvature operator @xmath23.\ ] ] the dynamical character of the system is further determined by a set of parameters characterising the relative role of dissipative , inertial and electromagnetic effects in addition to the driving by gradients of density and temperature . in particular , we specify collisionality @xmath24 , magnetic induction @xmath25 , electron inertia @xmath26 and ion inertia @xmath27 , where @xmath28 is the background gradient scale length , and @xmath29 is the sound speed .
we use a standard set of edge parameters given as @xmath30 , @xmath31 and @xmath32 .
the perpendicular scale length for the dalf model is set as @xmath33 to the density gradient length , and for the itg model as @xmath34 to the ion temperature gradient length , so that @xmath35 .
the itg model has otherwise identical parameters .
accurate experimental measurement of ion temperature profiles in the tokamak pedestal region are only recently being advanced and have previously been plagued by large uncertainties .
the present experimental knowledge of edge parameters can thus be seen commensurate with both the dalf and itg model cases used here .
the computational domain is set to @xmath36 nodes in units of the drift scale @xmath37 for @xmath38 and 16 nodes in one field line connection length ( @xmath39 ) in @xmath40 .
the dimensions are chosen to appropriately account for statistical isotropy in small scales in both perpendicular directions with satisfactory spectral overlap , and for an extended box size in the electron drift direction @xmath41 to achieve convergence . a box resolution down to @xmath12 guarantees inclusion of all scales necessary for the nonlinear dynamics that are esssential for the drift wave turbulence characteristics .
metric element @xmath42 , @xmath43 ( unshifted ) , normal curvature @xmath44 , geodesic curvature @xmath45 , magnetic field strength @xmath19 and local magnetic shear @xmath46 in a tokamak with elongation @xmath47 and triangularity @xmath48 ( dashed lines ) , compared to a configuration with @xmath49 and @xmath50 ( thin solid lines ) and to an actual asdex upgrade configuration ( bold solid lines).,width=321 ]
tokamak equilibria are computed by solving the grad - shafranov / lst - schlter equation with the code helena @xcite , implemented as described in ref .
a set of nested flux surfaces in straight field line hamada coordinates @xmath51 is obtained by specification of given radial profiles of pressure and rotational transform , and of the shape of the bounding last closed flux surface .
these hamada coordinates are then transformed into a field aligned system and re - scaled into local flux tube coordinates @xmath52 @xcite . global consistency in the parallel boundary condition for one connection length is maintained @xcite .
a transformation of the coordinate @xmath41 , which signifies the electron diamagnetic drift direction perpendicular to the magnetic field within flux surfaces , is applied in order to avoid grid deformation by local magnetic shear @xcite .
otherwise , grid cells are sheared strongly in @xmath41 direction particularly near the x - point region , with malign consequences on nonlinear dynamics , especially in the vorticity , that lead to a violation of the basic drift wave character and overestimate linear mhd - like dynamics .
the differential operators are then expressed in terms of the flux tube coordinates : the curvature operator becomes @xmath53 the perpendicular laplacian in flute mode ordering is @xmath54 and the parallel derivative is @xmath55 noting that the factor of @xmath56 in @xmath57 also depends on @xmath58 .
some metric coefficients @xmath59 that were obtained for elongation @xmath47 and @xmath49 with triangularity @xmath48 and @xmath50 are shown in fig . [
f : gxx ] .
increasing elongation @xmath7 specifically rises the local magnetic shear @xmath60 and reduces @xmath45 both in the upper and lower regions of the torus that correspond to flux tube coordinates @xmath61 . in the following , the configuration with @xmath47 and @xmath48 will be refered to as a `` simple circular torus '' ( sct ) .
local and global magnetic shear are in general known to have a damping influence on tokamak edge turbulence @xcite , whereas geodesic curvature acting through @xmath45 upon the axisymmetric component ( @xmath62 components , or `` modes '' ) maintains the coupling for a loss channel from zonal flow energy eventually to turbulent vortices @xcite .
both mechanisms help to reduce turbulent transport .
normal curvature @xmath44 on the other hand strengthens primarily the interchange forcing of the turbulence ( @xmath63 ) .
a series of tokamak equilibria is constructed for elongation @xmath64 and triangularity @xmath65 with equal profiles of pressure and rotational transform .
the flux tube is chosen at a radial position @xmath66 , where @xmath67 is the volume enclosed by the last closed flux surface .
the background density profile and rotational transform for the gem turbulence computations are linearised within the bounds of the radial computational domain .
we restrict our simulations for now to the closed field line region lying a few tens of ion gyroradii inside from the separatrix , thus avoiding complications that occur by a divergent metric and associated grid deformation when the last closed flux surface is approached in the vicinity of an x - point .
ultimately , the goal of edge turbulence simulations will be to combine sufficiently well resolved nonlocal drift wave computations with a representation of the realistic field line geometry crossing the separatrix to the bounded scrape - off layer region .
when the plasma shape is thus varied , we have a particular interest in the effects on fluctuation time and spatial scales , on the radial variation of flux surface averaged ( zonal ) flows , and on turbulent transport .
we first apply the ensemble of shaped tokamak equilibria to the dalf model .
the cross - field turbulent electron transport is determined by the flux surface average of the radial @xmath68 convection of the fluctuating density by @xmath69 , given in standard gyro - bohm normalisation to @xmath70 .
the transport flux @xmath71 is fluctuating in time , and for specifying a quantitative value to it the time average over a sufficiently long window , that covers all relevant frequency scales , is taken during the final phase of simulations after all initial linear transients and spin - up of flows and oscillations have reached saturation . as shown in fig .
[ f : kap - dalf ] , we find that @xmath72 is reduced to around a third when elongation is increased from a circular cross section to @xmath49 .
scaling @xmath72 against triangularity @xmath8 , we find in fig .
[ f : del - dalf ] that for @xmath47 the transport flux is independent from @xmath8 within the error bars given by the deviation of the fluctuating data from its time average . for higher elongation @xmath49 the transport @xmath72
is increased from the normalised value of @xmath73 to @xmath74 by increasing @xmath8 from @xmath75 to @xmath76 .
a good fit to the transport data in the dalf model within fluctuation error bars is obtained by @xmath77 where @xmath78 .
this reduction of transport by elongation can be related to three different causes : \(1 ) increase of local and global magnetic shear ( magnetic shear damping ) , ( 2 ) reduction of curvature drive ( interchange coupling ) , and ( 3 ) an enhancement of sheared zonal flows ( geodesic transfer ) .
dalf : dependence of turbulent electron transport @xmath72 ( in gyro - bohm units ) on elongation @xmath7 for triangularity @xmath48 , @xmath79 and @xmath76 ( from bottom to top).,width=283,height=207 ] dalf : dependence of turbulent electron transport @xmath72 ( in gyro - bohm units ) on triangularity @xmath8 for elongation @xmath80 , @xmath81 and @xmath82 ( from top to bottom).,width=283,height=207 ] in section vi we will present results showing that the effect of flow shear enhancement by a weakening of the geodesic transfer mechanism is insignificant for dalf parameters .
also for these resistive electromagnetic edge parameters , the interchange dynamics in the normal curvature forcing is acting more as a catalyst than as a drive mechanism for the turbulence , with only weak ballooning of the fluctuating quantities observed . on the other hand
, it has been established that local magnetic shear by flux surface deformation has a strong damping influence on tokamak edge turbulence : comparing the case of a simple circular torus ( with global shear only ) to a case where the local magetic shear structure is typical of a shaped tokamak ( but the curvature properties are those of a circular torus ) , a reduction of transport to around a half was found @xcite . moreover , the effective topological global shear ( defined , e.g. , in eq .
29 of ref .
@xcite ) is also increased by finite aspect ratio and elongation , but is not significantly affected by moderate triangularity ( @xmath83 or so ) .
it can thus be argued , that the major mechanism for reduction of transport by elongation in the dalf case is magnetic shear damping , with an additional ( slightly lesser ) influence by interchange drive reduction , and an only insignificant zonal flow enhancement .
we now include ion temperature gradient ( itg ) driven dynamics by allowing for a gradient ratio @xmath84 and using the full six - moment version of gem . in the itg model ,
the flux - surface averaged turbulent transport can be characterised by particle transport @xmath85 and by heat transport @xmath86 with convective component @xmath87 and conductive component @xmath88 , where @xmath89 is the gyro - averaged exb velocity and @xmath90 is its flr correction .
normalisation is to standard gyro - bohm units with the gradient scale length set to unity , and @xmath91 is the temperature ratio . when displaying heat and particle transport in relation to elongation and triangularity now for the itg model , we find in fig .
[ f : kap - itg ] a scaling of @xmath92 and @xmath93 .
the reduction by elongation is thus stronger than for the dalf model .
one evident mechanism for this increased reduction in the itg case is the weakening of the interchange drive of the itg instability acting via the normal curvature @xmath44 , which is slightly reduced by elongation in the ballooning region of the flux tube ( compare fig . [
f : gxx ] ) . in contrast to dalf dyanmics ,
the itg turbulence is strongly driven by `` unfavourable '' normal curvature , leading to a pronounced ballooning character of the fluctuations , as shown in fig .
[ f : balloon ] .
we find as an additional mechanism in the itg case a significant weakening of the geodesic transfer mechanism mediated by @xmath45 , which is discussed in the following section .
similar as in the dalf case , both heat and particle transport are slightly increasing with higher triangularity for an elliptical cross section ( @xmath49 ) in fig .
[ f : del - itg ] , while for small @xmath7 no influence from triangularity is observed within the fluctuation error bars . summarising the results of this section on the influence of flux surface shaping on turbulent transport for different edge parameters , we find that the main contribution to transport reduction in both the dalf and itg regimes is magnetic shear damping by elongation .
shear flow enhancement by the reduced geodesic transfer mechanisms is found to be relatively weak in the dalf case , but is becoming relevant in the itg model , as can bee seen from analysing the wave number and frequency spectra discussed in the next section . for fixed gradients ,
the local diffusivity is proportional to the transport and thus to the roughly inverse quadratic scaling with @xmath7 given above .
global transport across an entire magnetic flux surface can in principle be obtained by integration of the local transport coefficients , that takes the variation of surface area with elongation into account .
itg : dependence of turbulent particle transport @xmath94 and heat transport @xmath95 ( in gyro - bohm units ) on elongation @xmath7 for triangularity @xmath48.,width=283,height=207 ] itg : dependence of turbulent particle transport @xmath94 and heat transport @xmath95 ( in gyro - bohm units ) on triangularity @xmath8 for elongation @xmath47 and @xmath96.,width=283,height=207 ] itg : in the shaped configuration with @xmath49 the coupling quantity @xmath97 has a reduced squared amplitude compared to @xmath98 , indicating a weaker interchange ballooning character of the turbulence than in a circular torus.,width=283,height=207 ]
time traces are resolved down to @xmath99th of the drift frequency @xmath100 , and long computational runs are required to account for zonal flows with @xmath101 .
a typical number of time steps is @xmath102 .
geodesic acoustic oscillations ( gam ) may be distinctly detectable for some parameters and are expected for circular geometry around @xmath103 . in the dalf case for @xmath104
we have @xmath105 , and in the itg case for @xmath106 we have @xmath107 . in general ,
the gam frequency is determined by the geodesic curvature term @xmath45 and is influenced by flux surface deformation .
the geodesic curvature effect is the action of the curvature operator @xmath108 upon the axisymmetric component ( @xmath62 ) . in local flux tube
coordinates we find @xmath109 , and the factor @xmath110 in the geodesic curvature term does also accordingly scale the geodesic acoustic frequency : stronger elongation shifts the gam resonance frequency in the spectrum closer to the zero - frequency zonal flow .
the same argument , now applied to the normal curvature term @xmath44 , also accounts for a @xmath111 scaling of the ( interchange ) drift wave frequency .
for dalf parameters used in our computations , the gam peak does not protrude very distinctly but can still be identified on the flat top of the spectra . in fig .
[ f : freq - dalf ] the discrete fourier transform spectrum @xmath112 for a measurement of @xmath14 at a local point in the centre of the @xmath113-domain is shown : the spectrum below the gam resonance is mostly flat except for the distinct zonal flow peak at @xmath114 .
the higher frequency part shows a typical @xmath115 cascade structure , where the exponent @xmath116 changes around the drift frequency @xmath117 , which for our choice of edge parameters coincides also with the alfvn frequency .
one may identify three ranges in the spectrum : a flat top region between the zonal flows and gam frequencies , one intermediate range between the gam and drift wave frequencies with @xmath118 , and a cascade range between the drift wave and the dissipation times scale with @xmath119 .
as expected , both gam and drift wave frequencies are reduced by increasing elongation to lower values @xmath120 .
the whole spectrum is thus shifted to the left for @xmath49 in fig .
[ f : freq - dalf ] . using itg parameters , the gam peak is more clearly visible for the circular torus case but vanishes for elongation @xmath49 , as can be seen in fig .
[ f : freq - itg ] .
the whole spectrum experiences again a shift to lower frequencies for the elongated case .
the @xmath114 zonal flow component in the itg spectrum exhibits significantly higher amplitude for elongation @xmath49 than for the circular torus .
dalf : local frequency spectra @xmath121 for @xmath47 ( thin line ) and @xmath49 ( bold line ) , both for @xmath48 .
arrows indicate the drift wave and gam resonances for @xmath47.,width=283,height=204 ] itg : local frequency spectra @xmath122 for @xmath47 ( thin line ) and @xmath49 ( bold line ) , both for @xmath48 .
arrows indicate the drift wave and gam resonances for @xmath47.,width=283,height=204 ] in addition to computing the flux surface shapes with varying elongation and triangularity discussed above , helena is also used to obtain equilibria with an outer boundary defined by an experimentally reconstructed separatrix position .
in this way we obtain an equilibrium in the typical shape of an asdex upgrade ( aug ) @xcite lower single null divertor configuration ( as shown in fig .
[ f : aug ] ) .
the additional asymmetric shaping by the presence of an x - point in the lower part of the torus leads to an increase of local magnetic shear near @xmath123 and an associated local reduction of geodesic curvature .
the elongation and triangularity of this configuration are otherwise comparable to the kappa - delta model with @xmath124 and @xmath125 .
this aug configuration was first used in edge turbulence computations in ref .
@xcite , where the dependence of transport on @xmath126 was established : it was found that the onset of mhd ballooning mode turbulence is prevented for typical tokamak edge parameters due to the shaping effects in realistic geometry , and the nature of transport is still basically of the drift - alfvn wave character .
the aug model has , for the present flux tube position , properties relevant to the turbulence dynamics that are in some aspects in between those of up - down symmetric configurations with @xmath127 and @xmath82 .
transport in the aug model is for dalf parameters found to be reduced to @xmath128 compared to @xmath129 for the circular torus .
the nonlinear dalf growth rate @xmath130 is only slightly lower for the aug model with @xmath131 than for the circular torus with @xmath132 . for itg parameters ,
the particle transport is nearly identical in the aug and sct cases , both at @xmath133 .
the nonlinear itg growth rate is also similar , @xmath134 for aug and @xmath135 for the circular torus . a reduction of geodesic curvature by elongation and x - point shaping
might be expected to lead to a weakening of the geodesic transfer coupling mechanism for energy from zonal flows to gams @xcite .
this can be analysed by comparing the flux surface averaged vorticity @xmath136 for the configurations of the circular torus with the aug model .
the zonal vorticity @xmath137 , set in relation to the nonlinear drive rate @xmath138 , thus represents the radial shearing of zonal flows , which is considered responsible for the turbulent shear flow decorrelation and energetic damping of vortices @xcite .
concerning the geometric effect on this , we find different results for the dalf and itg parameters . in the dalf case , the cumulative amplitude of the components in the radial spectrum of flow shear @xmath137 is the same within fluctuation error bars for aug geometry as for the circular torus .
the lowest @xmath139 component in these spectra , which is found to be stronger in the aug case , makes up for only approximately 8% of the total vorticity .
the spectra are shown in fig .
[ f : zf - dalf ] .
flow shear in the dalf case thus is not significantly enhanced by flux surface shaping .
also the amplitude of zonal flows in the dalf frequency spectra in fig .
[ f : freq - dalf ] is found to be comparable for circular and shaped configurations .
the reduction of turbulent edge transport is in this case therefore mainly a result of the local magnetic shear effect @xcite . in the itg case , the flow shear rate deduced from the @xmath139 spectra of zonal vorticity is cleary higher in all components for the aug tokamak than for the circular torus , as can be seen in fig .
[ f : zf - itg ] .
the cumulative amplitude in the spectra is by 70% higher in aug geometry than in simple torus geometry .
flux surface shaping introduced by elongation and x - point shaping clearly enhances flow shear significantly for itg edge parameters . the zonal flow amplitude in fig .
[ f : freq - itg ] is also by a factor of three larger in the elongated configuration .
this shear flow enhancement contributes to the overall turbulent transport reduction by elongation , as has been discussed in section v , in addition to the effects of magnetic shear damping and the reduction of interchange drive .
the detailed relative importance of these three mechanisms may however change with varying plasma parameters .
dalf : @xmath139 spectra of flow shear rate in aug geometry ( bold line ) and for the simple circular torus ( thin line ) .
the cumulative flow shear amplitude in the spectra is similar for both cases.,width=283,height=207 ] itg : @xmath139 spectra of zonal vorticity related to nonlinear growth rate ( setting the flow shear ) in aug and sct geometry .
shear flows are cleary enhanced in the aug configuration in all components.,width=283,height=207 ]
we have discussed the effects of flux surface shaping on tokamak edge turbulence and flows .
elongation strongly reduces the turbulent transport , whereas triangularity was found to have only weak ( transport enhancing ) influence .
two regimes of edge plasma parameters were studied : an electromagnetic resistive drift wave ( dalf ) case , and the ( itg ) case with an additional ion temperature gradient of @xmath84 .
the scaling of local transport is roughly with the inverse square of elongation , showing a stronger reduction for itg parameters than for dalf parameters .
the itg result is consistent with earlier findings with the previous version of the gem model @xcite .
the mechanisms of flux surface shaping effects on turbulence have been discussed . the major cause for transport reduction by elongation is damping of turbulence by stronger local magnetic shear at the upper and lower parts of the torus on top of increased global magnetic shear .
the supervening reduction of the interchange drive by a lower normal curvature has an additional turbulence reduction effect in particular for itg parameters , where the interchange drive is slightly weakend . in the itg case a significant enhancement of zonal flow shear by elongation and x - point
shaping further contributes to transport reduction .
the shear flow effect is on the other hand found to be negligible for dalf parameters .
both the dalf and itg parameter regimes used in our computations seem consistent with the presently still poor experimental knowledge on edge ion temperature profiles , though this situation is rapidly improving @xcite .
it can be assumed that additional core plasma heating , e.g. by neutral beam injection ( nbi ) used in particular for obtaining the transition to an h - mode , is both increasing the ion temperature at the pedestal top and the ion temperature gradient in the overall edge pedestal region .
based on our results , the shaping of tokamak flux surfaces by elongation and by the additional presence of an x - point like in aug is expected to facilitate an enhancement of zonal flows when the role of the ion temperature gradient dynamics is pronounced by additional heating .
it can be expected that this enhancement of zonal flows found for itg parameters is even more pronounced for more strongly shaped flux surfaces nearer to the separatrix .
preliminary results suggest that such initial radially local flows have the ability to spread over wider regions of the pedestal in simulations with global profile evolution .
some l - h transition theories include zonal flows as a trigger mechanism to induce the transition to a sustained mean shear flow .
the enhancement of zonal flows by flux surface and x - point shaping in the presence of ion temperature gradient steepening may offer an additional explanation for the observation that h - mode access in tokamak experiments is facilitated by the presence of a divertor .
the possibility to obtain a confinement transition within first principle computations of edge turbulence will thus have to be studied with a code that includes full temperature dynamics , realistic flux surface geometry , global profile evolution , and a coupling of edge and sol regions including realistic sheath boundary conditions , while in addition it maintains sufficient grid resolution , grid deformation mitigation , and energy plus enstrophy conservation in the vortex / flow system .
this work was supported by the european commission under contract fu06-ct-2003 - 00332 . | shaping of magnetic flux surfaces is found to have a strong impact on turbulence and transport in tokamak edge plasmas .
a series of axisymmetric equilibria with varying elongation and triangularity , and a divertor configuration are implemented into a computational gyrofluid turbulence model .
the mechanisms of shaping effects on turbulence and flows are identified .
transport is mainly reduced by local magnetic shearing and an enhancement of zonal shear flows induced by elongation and x - point shaping . |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.